ASPRS approves edition 2 of the ASPRS Positional Accuracy Standards for Digital Geospatial Data

On Oct. 26, 2023, I participated in an American Society for Photogrammetry and Remote Sensing (ASPRS) Pacific Southwest Region Fall Technical webinar. The webinar provided an overview of the ASPRS Positional Accuracy Standards for Digital Geospatial Data (Edition 2, Version 1.0 – August 2023). The document can be downloaded here.

I also participated — virtually — in the Nov. 2, 2023, California Spatial Reference Center (CSRC) Coordinating Council fall meeting where Dr. Riadh Munjy, California State University, Fresno, discussed the revisions to the ASPRS Positional Accuracy Standards for Geospatial Data.

The most significant changes introduced in this second edition of the standards include:

1. Elimination of references to the 95% confidence level as an accuracy measure.
2. Relaxation of the accuracy requirement for ground control and checkpoints.
3. Consideration of survey checkpoint accuracy when computing final product accuracy.
4. Removal of the pass/fail requirement for Vegetated Vertical Accuracy (VVA) for lidar data.
5. Increase the minimum number of checkpoints required for product accuracy assessment from 20 to 30.
6. Limiting the maximum number of checkpoints for large projects to 120.
7. Introduction of a new term: three-dimensional positional accuracy.
8. Addition of Best Practices and Guidelines Addenda for:
   1. General Best Practices and Guidelines
   2. Field Surveying of Ground Control and Checkpoints
   3. Mapping with Photogrammetry
   4. Mapping with Lidar
   5. Mapping with UAS

As outlined above, Edition 2 contains Best Practices and Guidelines for (1) General Best Practices and Guidelines and (2) Field Surveying of Ground Control and Checkpoints. The three addenda listed in the table of contents: Mapping with Photogrammetry, Mapping with Lidar, and Mapping with UAS will be available for public comment later, and will be added to Edition 2, Version 2.0.

Dr. Abdullah informed me that these addenda are on track to be put out for public comments during December 2023, therefore he believes they will probably be published in January or February 2024. The box titled “Summary of Significant Changes in Edition 2” provides the changes with the reason and justification for each change. The document can be downloaded from ASPRS here.

One of the changes is to relax the accuracy requirement for ground control and checkpoints. At first glance, this seems like the wrong thing to do. However, after understanding the justification, the requirement for ground truth still needs to be at least twice as accurate as the product.

Both Dr. Abdullah and Dr. Munjy’s emphasized in their presentations that the current accuracy requirements for ground controls in photogrammetric work of four-times better than the produced products, and the checkpoint accuracy requirement is three-times better than the assessed product. This makes it difficult, if it is not impossible, to use RTK-based techniques for this type of surveying. This by itself is not the reason for the change. During Dr. Abdullah’s presentation, he provided the following reasons for the change:

1. “Experience taught us that the requirements of four-times and three-times adopted in edition 1 of the standards are excessive and too restrictive, partly due to the reason outlined in (b) below.
2. Today’s sensors, software, and processing methodology are more accurate and the room for errors in the product is diminishing, therefore we do not need a safety factor of 3 or 4 to obtain accurate products.
3. Increasing demand for higher accuracy geospatial products.”

The new standards now factor in the accuracy of the survey checkpoints when determining the accuracy of the product. During Dr. Abdullah’s presentation, he provided the following reason for the change, “As we are producing more accurate products, errors in surveying techniques of the checkpoints used to assess product accuracy, although small, can no longer be neglected and it should be represented in computing the product accuracy.” He also highlighted that, “As product accuracy increases, the impact of error in checkpoints on the computed product accuracy increases.” The document provides equations used to compute the values. See below.

A very significant change, in my opinion, is the removal of the standards for Vegetated Vertical Accuracy (VVA) for lidar data. See below.

I am not sure I agree with the reasoning, but I understand why it was done. GNSS-based surveys do not perform well in vegetated areas, and this is the technology used to validate the non-vegetated vertical accuracies (NVA). That said, there are non-GNSS technologies — sometimes denoted as traditional surveying methods — that could be used to validate VVA, so this seems like an elimination of a requirement based on the limitation of a particular technology.

Traditional surveying methods that use geodetic levels, theodolites, and total stations to measure distances, angles, and heights are still used by surveyors to perform certain projects. Since there are other surveying methods that could be used for evaluating the VVA, it does not seem like a valid reason for a change.

The ASPRS standards does state that, “for projects where vegetated terrain is dominant, the data producer and the client may agree on an acceptable threshold for the VVA.” Therefore, the client can require the surveyor to meet a specific accuracy level for vegetated areas. I am sure this was discussed during the working meeting, so I leave it to the experts to make the appropriate decisions and recommendations.

Finally, it should be noted that, as discussed above, the new ASPRS standards eliminated the reference to the 95% confidence level as an accuracy measure. The document provides the following statement about the National Standard for Spatial Data Accuracy (NSSDA):

“The National Standard for Spatial Data Accuracy (NSSDA) documents the equations for the computation of RMSE_X, RMSE_Y, RMSE_R and RMSE_Z, as well as horizontal (radial) and vertical accuracies at the 95% confidence levels — Accuracy_R and Accuracy_Z, respectively. These statistics assume that errors approximate a normal error distribution and that the mean error is small relative to the target accuracy. The ASPRS Positional Accuracy Standards for Digital Geospatial Data reporting methodology is based on RMSE alone, and thus differs from the NSSDA reporting methodology. Additionally, these Standards include error inherited from ground control and checkpoints in the computed final product accuracy.”

Appendix D of the ASPRS document provides the equations with an example for computing the accuracy statistics. The document also has a section with examples for users who wish to relate the ASPRS 2023 Standards to the FGDC National Standard for Spatial Data Accuracy (NSSDA).

Dr. Munjy ended his presentation at the CSRS 2023 fall meeting with the following statements:

“ASPRS Accuracy Standards 2023 have become more aligned with science and statistical theory,” and “These Standards are intended to be a living document which can be updated in future editions to reflect changing technologies and user needs.”

I would encourage all users to download the document to better understand the changes and reasons for the changes. It can be downloaded here.