Hey, my question for today is a general one, regarding PPK 
So I understand you can use either a.) a standalone rover or b.) a base station AND a rover for the GCP stakeout for the PPK processing.
Assuming you are using the same manufacturer’s product in both cases, what kind of accuracy improvement can you achieve with using both?
I would like to achieve cm accuracy, can I do it with just one rover? I know that depends on the rover, and I’m open for suggestions.
are you looking to correct gps data collected by your drone or getting accurate GCP data?
If you have a suitable station nearby where you can get the corrections data from, and your rover has reasonable accuracy to begin with, you can normally get centimeter-class at least.
This blog is an incredible resouce on RTKlib, but also GPS hardware and processing in general:
let’s explain it a bit further: I have my eyes on the Emlid reach rs+ or the Ardusimple RTK Calibrated Surveyor Kit.
How I have imagined the workflow:
So do I need a base station + a rover for cm accuracy or is it enough to stake out the GCP’s alone if I don’t plan using NTRIP?
To perform high accuracy Post Processed Kinematic (PPK) measurements generally you require a base station nearby to collect GNSS data on a known highly accurate control station while the rover travels from GCP to GCP in kinematic mode stopping momentarily to take an average of measurements in a “stop and go” sequence. However, there are ways to achieve similar accuracies with only one rover without the need for a base station if you have a GNSS receiver that is capable of reading NTRIP corrections in real time. Where I live NYSDOT has an NTRP service that is free to use here in New York State. I have performed some accuracy tests on their NTRP system utilizing my Emlid M2 GNSS receiver occupying several New York High Accuracy Reference Network (NYHARN) control stations located around the state. I performed these tests back in 2020 and my demanding retirement schedule has prevented me from posting these findings in Emlid’s form. I plan to do that soon. The link below references my testing of static observations with the same receiver on their forum. The table below illustrates the accuracies I achieved while performing two independent 2 minute occupations separated by a minimum of 20 minutes on each control point. You can draw your own conclusion as to if these accuracies are good enough for your requirements.
Difference between average of 2 independant occupations vs. published value
| ∆ Easting (m) | ∆ Northing (m) | ∆ Elevation (m) | |
|---|---|---|---|
| Ben1 | 0.018 | 0.006 | 0.004 |
| Cobleskill | 0.011 | 0.004 | -0.011 |
| Corinth | 0.003 | -0.005 | -0.016 |
| E 184 Reset | -0.011 | -0.004 | 0.011 |
| Ft Miller | 0.001 | 0.007 | -0.032 |
| Martin | -0.001 | 0.008 | -0.031 |
| N296 | -0.004 | 0.007 | -0.055 |
| New Scot | 0.003 | -0.016 | 0.007 |
| Utica | -0.002 | 0.001 | -0.020 |
| Wright | 0.005 | -0.002 | -0.059 |
| Avg of All Values | 0.002 | 0.001 | -0.020 |
Beautiful stuff! We’re fortunate to have someone with your expertise among our Community!
NYSNET CORs is pretty awesome for free, isn’t it?
I actually worked in the Survey Section and was part of the Geodetic GPS team that established the original NYHARN with the cooperation of the National Geodetic Survey (NGS) back in the mid 1990’s.
Thank you!
What can you tell us about corrections by NTRIP? would using rtk2go caster for corrections be a viable solution?
Thanks
I’m not familiar with RTK2Go but it is probably similar to NYSDOT’s system. NYSDOT’s NTRIP reference station corrections are broadcast on NAD83 (2011) . You will need and internet connection through your cell phone connected to the GNSS receiver to receive NTRIP corrections. Here is the FAQ page from NYSDOT that explains some of this. There is a lot of info on this on the internet. I know that the current NYSDOT survey crews use NTRIP to locate photo control for highway design projects. The system is highly reliable.
I read something interesting just now while comparing the Emlid RS+ with the RS2:
Accuracy in PPP mode for the RS2: “Centimeter accuracy”
Am I reading it correctly, in PPP mode for my use case I would only need one single RS2 GNSS receiver, no base station or NTRIP correction?
https://gssc.esa.int/navipedia/index.php/Precise_Point_Positioning
You would have to stay static for a very long time at each target to achieve highly accurate results. You will be way more productive with the RS2 than the RS+ because the RS+ is single frequency. If cost is the issue then You could get an M2 setup like I described in an earlier link in this thread for about $800. Here is a link to someone in the Emlid forum that tested the RS2 in PPP.
Copied from the site:
RTK2go® is a community NTRIP Caster created to allow you to publish your GNSS correction streams for others to use with their NTRIP Clients. It is built using the same Pro edition of the SNIP Caster you can find on the use-SNIP.com site.
I would test that service using a scenario like I did with the NYSDOT NTRIP service. Find a nearby high accuracy control point to occupy briefly and see what accuracy you achieve.
for what I do something like this will do me just fine. https://www.ardusimple.com/product/simplertk2blite-basic-starter-kit-ip67/ Plan to build it here in the near future and use NTRIP for correction. Unless of course I am out in the middle of nowhere with little to no broadband connection, I may have to build another as a base.
That looks like a great project. What are you going to use for a data collector? I don’t see anything about storing RINEX or UBX files. You will need the raw data if using with a base when you do not have an internet connection. And somehow collect stop and go kinematic data when locating targets with the NTRIP service…
Plan is to use my cell phone with the mapit gis app. that app has the ability to connect to NTRIP correction as well as connect to a GNSS receiver through usb or bluetooth. through mapit I will record data. This is all a learning experience for me and this forum is helping tremendously.
If I am ever in need of a base station I will have to do some more research.
I will be interested to see how your project turns out. Your MapIt solution should work.
will follow up when I get to that point. Thanks for your input sir.
So I received my GNSS receiver from Ardusimple. I purchased a rover from Amazon for a whopping $44, to which I was surprised at the quality at this price given that other rover poles I found on the internet were closer to $100 or more. I made a small platform out of sheet metal to mount the antenna to which is supposed to help in the reception for the multiband antenna. The unit I bought from Ardusimple is a preconfigured unit that employs bluetooth for connection to a cell phone or tablet.
I am currently using the SW maps apps that is free for android devices on google play. It connects to the receiver quickly and has a NTRIP client built in. It records data points with averaging and export the data in all kinds of formats. csv,geojson,kmx,kml,excel xls, or ods if you use open office and geopackage.
as grand total I have probably spent around $400.
So far i have only tried this around my house which is surrounded by trees and have been around 400mm meter accuracy. I am hoping for more, but may need to get a better antenna to achieve that. However I plan to test the accuracy in more wide open areas with better satellite reception to see how it fares.
I am curious however if I were to use a different NTRIP caster closer to my location if that would help with accuracy? As tested I used the closest NTRIP caster to me which is about 250km away.
