I am interested to get the opinion of the community on the feasibility of GNSS in archaeological excavations. GNSS has been used quite successfully in archaeological surveying and in extracting GCPs for drone surveying. But does it have enough accuracy and consistency to potentially replace the Total Station? For many excavations (my experience is in Mediterranean archaeology), sub-centimeter accuracy may be required. But for others (and this may be a majority), an error of 1-3 cm in horizontal and vertical would probably be acceptable. I outline a scenario below and list some potential advantages and disadvantages.
Context: large open sky coastal site with almost no trees or urban structures. Some low vegetation. Minor multipath concerns from a chain link fence situated a few meters from one of the excavation areas. Elsewhere, telephone wires ca. 50 m further outside of the area of interest.
Setup: Base will not be set over a known datum, but it will be connected to a nearby CORS network no more than 200 m from all intended excavation areas. A Fix will be obtained and then the base position averaged for 10 min. Most if not all excavation areas will have an unobstructed line of site to the Base. Rover units will be connected to the Base via LoRa. Multiple Rover units will be used, ca. one for each of the excavation areas. iPads with ReachView 3 will be connected to the Rovers’ hotspot. Collection points will be averaged for 30 seconds.
Advantages: (1) quicker setup than a Total Station over a datum; (2) no need to have personnel at the Base throughout the day, as would be the case with a (non-robotic) Total Station (this also saves personnel costs during a season); (3) efficiency in the speed of data collection, e.g. each excavation area takes points whenever needed with the Rover and there is no time wasted waiting for the Total Station to finish measurements in another area; (4) possibility to collect data where there is not a perfect line of site to the Base, unlike a Total Station which could not be used in the same situation.
Disadvantages: (1) satellite coverage and atmospheric conditions could potentially affect the measurements from day to day; (2) vertical accuracy, which is critical in archaeological excavations, will have a greater error factor than a Total Station would provide.
I’d recommend using the RS2 (dual freq). The only other issue is the baseline length from the nearest CORS.
The RS2 baseline specs are 60km for RTK and 100km for post processing. An alternative is to log the base position for minimum of 8 hours and the submit to AUSPOS or NRCAN for processing using precise emphemerides. You can then convert to your local projection/datum/geoid. Upload the converted base point coords to RV3 and select your system for the base and then you are good to go.
Normally you can expect 2-3cm which should be plenty good for your use case. I like the 10min average but if you can connect to CORS then you just use the rover to set a point and then manually enter the coords into the base and run the local caster. Relative accuracy would be fine if you are not concerned with cm-level global positioning.
Another option which we use fairly often is to shoot a base point at a more acceptable distance from the CORS and continue setting additional points as you traverse to the site. If you are truly 200km then there has to be some sort of monument closer to the site that you can traverse from.
One more option is a PPP base point but I’m not familiar with the process. @EBE111057 would be a good source for this
I don’t think I’ve ever been within 5 km of a CORS so 200 m messes with my mind. If it really is 200 m then just run the rover in the drone off of that. That way all of your positioning is from the same datum.
Hi Lucas. I think you have more or less presented the advantages and disadvantages pretty well. And, I think you may have also answered your own question actually.
Even though you have some projects where accuracy within 3cm would be acceptable, when using a gnss receiver you will never know absolutely for sure how far out it is unless you are checking a known point, or double checking it with the total station or direct measurement. In building and infrastructure, a gnss receiver is used all the time to get very close. But when it has to be a good as current tech allows, then the gnss receiver is not the tool.
So, if your situation is that knowing that the collected coordinate is “probably” within 3cm, then the gnss receiver can probably do it for you if conditions are good. However, if you have to know for sure that the point is sub-centimeter, or even <3cm, then the gnss receiver will have to be backed up in some way which might negate the time savings elsewhere.
This is a good point I should have mentioned. It is advisable to set at least two points on opposite ends of the area so that you have good check-in locations. This also gives you a couple of options as to where you can setup your base. If you can run on NTRIP then set two good points that are safe and be checked in to before each session.
Hi Lucas! I have nothing to add to the comments above.
A total station will be the right choice for cases requiring a subcentimeter accuracy. On other occasions, the GNSS receiver should work just fine.
You can use Reach RS+ or Reach RS2. Both models are applicable for your tasks. However, if you have working areas with a more complex environment than described, I suggest you look for Reach RS2. As a multi-band receiver, it’s robust to multi-path effects and partially blocked sky conditions.
If it really is 200 m then just run the rover in the drone off of that. That way all of your positioning is from the same datum.