Base coordinates in CSV-file from a rover with NTRIP corrections


I am quite new on using Emlid. Right now, I use an Emlid rover with NTRIP/VRS corrections and got it working with a GPR device. I use the Emlid flow app as an additional source for measuring some refernce points. When I look att my data, there are some coordinates for BASE which matches well with the first point that I have measured but it has a higher elevation. What does that mean, as I do not have base station but get corrections from a VRS service. How does this sound to you experts there?


The VRS stands for Virtual reference station. You are still getting correction data from each fixed cors station, however the information (I would assume) is revised and run through an equation to reduce baseline error. The VRS is calculated typically as close to your initial acquisition location as possible. The elevation difference could be caused by a bunch of different factors. Do you have the correct antenna heights input? Do you have the correct geiod selected? does the Cors station actually transmit Elevation Data? Are you looking at the difference between ellipsoid and orthometric heights? You now have enough information and words to complete some research, look at your settings and values you are comparing, and come to a conclusion. The best way to test your settings are accurate are to go to a known bench mark, start your project, get a fixed signal from the VRS and then take a shot on the bench mark. IF you have your information entered correctly the results (Point information collected) should match or be within 1 cm or less of the bench mark. Good luck!


Hi @Omid,

I can second Aaron’s words. VRS generates a virtual base somewhere close to your location, based on the data from the physical bases and your position. It still has exact coordinates and height.

I also have some questions about the elevation difference that you notice. What values do you compare? The height of the virtual base and of the first measured point? And what is the difference between them?


Thanks Julia, yes I compare coordinates of the first measured point and the base in my data. The height difference is about 4 m.

I should mention that I noticed later that the configuration for receiving base corrections weren’t so optimal as it was using single averaging in 5 seconds! I changed these to fix averaging in 40 seconds and I am going to test it again on a known point.


Oh, I see. Feel free to share the results of your tests on a known point here.

1 Like

@julia.shestakova But I think, it should be normal that the virtual base height is higher than the height of the measured points. Isn’t it?

@fullersports1016 Thanks for your explanations, I think I have a better understanding now. But will do more research on this matter.

1 Like

What is the elevation difference?
This sounds like 1 of 4 things:

  1. Incorrect vertical coordinate at rover
  2. Antenna height bust at rover
  3. Different vertical datum than VRS station
  4. Wrong geoid applied
1 Like


If you are using NTRIP your base configuration of averaging time should not matter. If you are using a local base see below.

There will be a difference in height if you are using a local base, you should have a direct shot on a hub or control point with a known elevation. In the event that is not an option due to distance, or something, than take a static or rapid static observation post process with opus or similar then set your base up with the computed coordinates.


I’d not compare the coordinates of the virtual base with the measured points at all. The place where the virtual base is generated depends on the algorithm of the exact service. The main thing is that it should be somewhere close to your position.

Plus, the VRS service uses the coordinates of the receiver in Single, which is several meters accurate. So I think it’s fine if the VRS base is created several meters higher or lower than your receiver.

At the same time, if you notice a mismatch in measurements, please let me know. In such a case, it indeed may be a question of configuration and coordinate systems. So we’ll need to thoroughly check it all.


@julia.shestakova Thanks, you are right and I fully understand the concept now.

1 Like

This topic was automatically closed 100 days after the last reply. New replies are no longer allowed.