I am using ArcGIS 10.3.1 for Desktop. I have a raw .las file that I have brought into an ArcGIS .lasd file format.

The raw data has several anomalous points; i.e. the maximum elevation for any point should only be 100 meters but there are perhaps 30 points with an elevation above 1000 meters.

I would like to delete these anomalous points.

The points are unclassified except for "ground". All others, including the anomalous ones, are "unclassified".

There doesn't seem to be a way to select these, or filter them out and then write a new .las / .lasd file that has eliminated them.

  • why not use a LAS layer?
    – radouxju
    Sep 12, 2015 at 19:46
  • Thanks but I am going to be doing further work with this file - trying to get tree canopy - and so I really need to get rid of these points all together.
    – Lynn
    Sep 12, 2015 at 23:38
  • 1
    I have found LASTools, rapidlasso.com which includes a custom toolbox....the lasheight tool has enabled me to delete the points.
    – Lynn
    Sep 16, 2015 at 1:09
  • 1
    great, I suggest that you add this as an answer to your own question (with a few additionnal details on how to use the lasheight tool etc) and you'll get my upvote.
    – radouxju
    Sep 16, 2015 at 6:41

2 Answers 2


As you cited, lasheight from LAStools can do the trick (used from ArcGIS). It says:

Another alternative is to use the computed height to eliminate points with a particular ground height above or below a threshold ... .


lasheight -i *.las -odix _no_outlier -olas -drop_above 100.0

The above example will make a batch call (i.e., with multiple input files), will keep points with height up to 100 units (e.g. 100 meters) above the ground, and will append the suffix name 'no_outlier' in all the output las files.

So, you would need to identify where is the gap between the outliers from the other points, and set the correct threshold value for -drop_above.

Another option is the native tool from ArcGIS Desktop Locate Outliers:

Identifies anomalous elevation measurements from ... LAS datasets that exceed a defined range of elevation values ... .

You would set the tool with the Apply Hard Limit methot and the Absolute Z Maximum parameter equal to 100 meters.


Might be worth a thought: If the undesired points are significantly(or just noticeably) higher than the rest of the data, you can convert to .txt, open in spreadsheet editor, sort z-value column by descending, manually remove all points higher than desired data height, and then convert back to .las

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.