Collection time for topo data?

What are current industry standards for point collection time for topo data?

Looking at my state DOT standards, they specify a minimum of 6 epochs per observation. I believe those standards were written when only GPS and GLONASS were available, now our CORS system also broadcasts BAIDU and GALILEO

I’m used to spending more like 30 seconds of averaging on each point. Is two seconds enough?

Pretty hard to say it’s standards but I co-authored something on that 10 years ago for the national geographic institute - page 24, sorry it’s in French :sweat_smile:

There we recommended to stay 3 to 5 seconds per point, and 180 seconds on important points, such as benchmarks. For this latter kind of points, we also recommended to measure them again after 20 - 30 minutes.

So, 30 seconds on each point as you do is probably overkill if they are not benchmarks.


In the old days where we used L1 only GPS receivers with stop’n go method the recommended time for topo points was 10 sec.
Today with RTK 2-5 seconds per point is enough. If collecting more important points like stations or benchmarks then 30 seconds is better. Of course you should repeat the important measurements after 2-3 hours and average them to be more precise.


This is “standard” advice. But, I’ve always wondered if it is valid. If 2 (or more) observations are made on a point and averaged, the accuracy of the point has the potential to get better, but also the potential to get worse, correct? If one of the observations is out for some reason and you average it in. Doesn’t the averaged point accuracy degrade because of it?

What vetting of an observation do you do before it is added to the average?


1 Like

Indeed. This way you can actually check if one of the two collections is off. But you will need a third one to know which one :slight_smile:

If they are within 2-3 centimeters, no problem. The average will enhance the absolute accuracy of the point.

1 Like

As Florian said, a third measurement must be done to ensure the results.
The difference will be mainly in elevation. I have done extensive tests with my gear and collect rtk data in 10 seconds interval for about 3-5 hours. The XY accuracy was about 2-3 cm and the Z accuracy was about ±10cm.
All the tests were done with small baseline ~1km and various receivers (not emlid).


Hi @b_airben,

I agree with the suggestions of the guys below about 2-5 and 3-5 seconds per point in RTK. But I do have a few comments.

If you work with Reach RS2+ or Reach RX with the 5 Hz update rate, you get 5 epochs per second. So the 30 seconds of averaging is already 150 epochs.

I’d only recommend such averaging for repeatable measurements on geodetic reference frame points or traverse points. For survey points the mentioned periods are sufficient.

1 Like