Base station in average float and rover in fix

Hello,

I want to make some measurements in crops of weed locations. I am in Australia and I am quite new on the survey topic. Here is my setup:

  • RS+ as a base station connected to a NTRIP provider.
  • RS+ as a rover connected to the base using LoRa.

I have read that Baidu is better in Australia in terms of coverage. Unfortunately the base station seems to struggle to reach a Fix Solution. It might be linked to the fact that the NTRIP station has a baseline ~30km from the RS+ base. Here is the Ntrip provider Australia Ntrip. Is it possible to reach a Fix solution in this configuration ? Due to the distances it’s not possible to use a survey marker nor a closer Ntrip station.

So I usually use Average Float with a accumulation of 5 min. The rover in this configuration will reach a fix solution. Does it means that it’s a centimeter precision ?

I get an RMS of:

  • N = 0.011
  • E = 0.010
  • U = 0.010

Does it means that we have an error in a sphere of 20cm of diameter ? I didn’t managed to find a clear explanation on RMS & accuracy.

What would be also the correct observation time ? for now it’s 5s (default value).

Cheers

3 Likes

Yep, for a single frequency receiver like the RS+, 30 km baseline is indeed to much. Emlid officially says 10 km, I would advise closer to 6 km.

Usually, yes

No more like a 20 mm sphere of precision.
Notice that this is the precision relative to the base, not necessarily the absolute accuracy.

5 Likes

Thank you for your feedback !

Unfortunately that’s not possible. The closest is 26km.

How do I know the absolute accuracy ? I don’t necessary require centimeter precision if I can locate the weeds in a circle on the ground of 20cm of diameter I am ok with that.

1 Like

You need to get an RS2. That would solve your baseline problem. Regarding absolute accuracy, I would imagine you are good regarding your 20cm accuracy requirement. It depends on whether your base gets a good enough fixed or float position from your NTRIP provider (at that longer baseline). Why not just do stop and go with PPK with AUSPOS?

2 Likes

Then PPP is probably the better way. 26 km is a bit over the limit for single frequency baseline post processing. 19-20 km works better.

You don’t really know for certain, actually. You have to trust the technology and the specs. However, if you have a fix, you can usually count on than the point is corresponding to error-specification mentioned in the RS+ specification.
Anyhow, 20 cm is no problem, when you have a fix going.

2 Likes

Hi,

Welcome to our community forum!

I agree with some of the guys’ suggestions but want to share several points that can be of help.

Even the 26 kilometers is a too large baseline for Reach RS+ in RTK. But PPP is not the solution since Reach RS+ is a single-band. Australian PPP service AUSPOS doesn’t accept the raw data logs from the single-band receivers.

I assume that you can use two ways to obtain absolute accuracy here. The first is to place the base over a known point. These points are usually benchmarks, and you need to know their geographic coordinates (lat/long) in GDA2020 or GDA94. This approach doesn’t suit you if you don’t have them in your survey area.

The second one is the PPK technique. You can obtain the centimeter-level absolute accuracy after post-processing instead of getting an RTK solution over an NTRIP. You need raw data logs from the NTRIP base and the rover. You can ask an NTRIP provider for the base log, the rover log you’re able to collect on your own. You can post-process logs in our free software Emlid Studio.

This guide explains how to log data for PPK. You can check the common description in this section. Post-processing is explained here.

4 Likes

NRCAN should cover world-wide? Sure, it’s single-frequency, but still better absolute accuracy than a single frequency float over 26 km.

5 Likes

Hi Christian,

Yes, it covers. But it outputs coordinates in NAD83, while @austriker is in Australia. I assume that he doesn’t need this datum. I also think that the PPK technique is appropriate because the baseline is up to 30 kilometers with single-band Reach RS+. It requires recording raw data and post-processing but provides you with centimeter-level absolute accuracy.

1 Like

You get a choice of NAD83 or ITRF14 :smiley:

In my experience with the RS+, to reliable get a trustworthy fix, the baseline need to be under 19-21 km.

3 Likes

You’re absolutely right, I missed that fact :smile:

I believe it, but our specs allow doing PPK on baseline up to 30 kilometers. I assume that it’s possible in a very good conditions.

1 Like

I read that you need an observation time of 1 hour. Is for the base only ? Or should it be also for the rover ?

Do I need the Reach RS+ as a Base connected to the NTRIP or just NTRIP + the rover will do the job ?

Do you have a linux version ? Is there an alternative if it’s not the case ?

Typically, PPP provides cm level with dual frequeny receivers.
With a single frequency receiver typically PPP will provide half meter accuracy.
That being said, in my opinion PPP is a technique for dual fr. receivers!

1 Like

This observation is for obtaining the base’s absolute accuracy. Then, you can input the base coordinates manually and work in RTK via LoRa with the rover.

This approach is used to achieve absolute accuracy of the base position. It would be much easier to obtain FIX in RTK via LoRa on a shorter baseline than doing PPK only with the rover on a longer baseline up to 30 kilometers.

We have only Windows and macOS versions. I suppose that there are some versions of RTKLib that may be an option, but I can hardly help in installing them.

1 Like

Not sure where in AU you are, I am in Tasmania, you should have access to the Land Survey Control Marks, every state and territory maintain and publish data, in a short form it will look something like this
Site 520/100
Horizontal Information for 520/100
South Latitude East Longitude Zone Easting Northing
43 20 29.8024 146 57 08.2146 55 496132.188 5201247.556
Datum Geocentric Datum of Aust 2020
Pos. Uncertainty 0.15
Pos. Uncertainty Method Estimated

As you can see we have geographic coordinates convert to dec degrees abd put that into your base and then do NTRIP over LoRa. You can always build your own network to span a distance.
If you are in QLD get the Android app BENCHMRK. Pay attention to the Zone in AU we use GDAy2020

2 Likes

You need to get an RS2. That would solve your baseline problem. Regarding absolute accuracy, I would imagine you are good regarding your 20cm accuracy requirement. It depends on whether your base gets a good enough fixed or float position from your NTRIP provider (at that longer baseline). Why not just do stop and go with PPK with AUSPOS?