Effects of the Pole on Survey Accuracy?

Hello folks,
we are currently testing the accuracy of the Emlid products we use for our UAV surveys and reference data collection.

In one of the tests we mounted a RS2+ unit on a 2.1 meter carbon survey pole wit attached bipod to keep it stable during point measurements. The total distance from the antenna is reported as 2.234m (Adjusting for the Antenna of the RS2+)

The carbon pole has a bubble level with an accuracy of 40’. If we do the math: sin((40/60)°) * 2.234m = 26mm. So we could have an lateral error of 26mm in our measurement which is not neglibile.

Now we could order a more precise bubble level or get a smaller pole to reduce this error. However i was wondering whether we actually need to be that precise or if the Emlid reach units already account for this.

We set the pole length in Emlid Reach and the units have an IMU which can measure the level.
Do the Emlied Reach units correct for off-level placement ? And if yes do we need be accurate with leveling at all? If we can just eyeball the pole to be approximately level this can speed up point acquisition a lot.

1 Like

According to this discussion the Emlid units don’t include tilt into the calculation of the surveyed point. That means we will add some error by off nadir.

In the marketing materials it looks like this is not an issue and measuring points without a bipod is possible. In construction it’s the resulting accuracy might be sufficient though?

Emlid offers also a survey pole for the RS2+ unit. What is the accuracy of the bubble level?

The best bubble level I cloud find has 10’ (arcminutes) accuracy (6mm lateral error @ 2.234m). Does it make sense to use very short pole even if it means the the antenna has a less optimal position?

1 Like

2.6cm isn’t good? Who makes a 2.1m pole? Just so I can make sure I never accidentally buy one. We use 40’ vials all the time and hit within our 0.10ft tolerance easily every time. If you need more accuracy you need a robot.


Small hint : if you use a bipod during measurements, you can record the point twice by rotating the pole 180° between both takes and this way, the average cancels the plane error.


Can you explain that a little further? How is that going to help Z?

It would just help XY. Z wouldn’t/shouldn’t be affected by rotating 180. I am seeing better than a tenth vertically whether using VRS, caster, or lora if baselines aren’t too long.

1 Like

I agree with Michael, if you’re looking for more accuracy you would need a robot.
But here are some other possible solutions to gain more accurate placement of your GNSS receiver.
The Marksman

Typically used with prisms, though I seen it demonstrated using it with GNSS Receivers. Pictured with no extension pole, but I seen examples with various size poles for placing the receiver higher.

Or Javad makes a collapsible Monopod for their units that includes a dual 8min/40min bubble vial. Only thing is that it uses a 1/4-20 thread, so you would need an adapter to fit the RS2 5/8 thread. They also sell the double vial by itself.

Or you can visit Baseline Equipment CO. for various SECO Vial and Vial Assembly Replacements from 8min to 40min units

In practice, this is less, if you make it habit to always have the bubble in the very center of the circle.

Remember that including an IMU in these calculations also introduces yet another error-source, that contributes with its own error budget. So mechanical levelling is to be preferred, if possible.

1 Like

This does not help Z indeed.