Surveying old quarry

[panyasam]

Hi,

I'm working in the field of ecovillage design and would like some advice on drone mapping. I often do surveying with tape measures and a dumpy level but a project I'm working on at the moment has an old quarry that I'd like to map and it's difficult to get in there. It's not so big, about 100m (300-ish ft) across and maybe 20m (60-ish ft) deep. My only concern is relative accuracy for width, depth and slope angle. With these measurements I can design the required earthworks. My question is: 'do I need RTK for GCP, or will a drone alone give me enough accuracy -- say, within a foot or two.'

Any advice will be much appreciated.

Sam

 

[Meta4]

My question is: 'do I need RTK for GCP, or will a drone alone give me enough accuracy -- say, within a foot or two.'

You should be able to get better accuracy than you are wanting with GCP and no need for RTK.

 

[Dirt Pilot]

For one mapping job hire a drone pilot!

If you want to DIY then buy a drone, Phantom 4 ProV2 and accessories for $3000, or a Phantom RTK for $7000. Get your FAA Remote Pilot's License after studying for a month ($150). Then spend time and money on drone mapping software such as Drone Deploy = $300/mo + $50 per map with GCP's.

I would be glad to give further advice on mapping. I'd help you map it, if the quary was in the SF Bay Area.

 

[R Martin]

My question is: 'do I need RTK for GCP, or will a drone alone give me enough accuracy -- say, within a foot or two.'

Probably not. The last time we tested the accuracy of the on-board GPS unit it was within about five feet and shifted. The only way to pull it off without an RTK/PPK setup is to add GCPs liberally and shoot them in with a reliable GPS unit yourself -or- pay someone to do it for you (the GCPs). I normally correct via an RTK network but have a base station that integrates into our operations if the wifi connections is unreliable. I still utilize control points for no other reason than peace of mind in case everything dies. At least the flight doesn't have to be repeated and the data collected can be used and is accurate down to 0.5 px/in GSD (or slightly better in our case)

 

[panyasam]

For one mapping job hire a drone pilot!

If you want to DIY then buy a drone, Phantom 4 ProV2 and accessories for $3000, or a Phantom RTK for $7000. Get your FAA Remote Pilot's License after studying for a month ($150). Then spend time and money on drone mapping software such as Drone Deploy = $300/mo + $50 per map with GCP's.

I would be glad to give further advice on mapping. I'd help you map it, if the quary was in the SF Bay Area.View attachment 119951

Thanks for the response, and the offer, very kind. The quarry is in Thailand, though.

I tried to hire someone to map it and they were asking for more than twice the price of the drone just to produce one map! My client is pretty keen to buy a drone and I'm gonna outsource the data processing, just need to gather the data. Figured it couldn't be that difficult...

 

[panyasam]

Probably not. The last time we tested the accuracy of the on-board GPS unit it was within about five feet and shifted. The only way to pull it off without an RTK/PPK setup is to add GCPs liberally and shoot them in with a reliable GPS unit yourself -or- pay someone to do it for you (the GCPs). I normally correct via an RTK network but have a base station that integrates into our operations if the wifi connections is unreliable. I still utilize control points for no other reason than peace of mind in case everything dies. At least the flight doesn't have to be repeated and the data collected can be used and is accurate down to 0.5 px/in GSD (or slightly better in our case)

Thanks for the response.

My understanding was that the GPS gave absolute accuracy. I really only need relative accuracy. Can I not just use relative GCPs and physically (tape measure and dumpy level) plot their coordinates relative to each other?

 

[R Martin]

Thanks for the response.

My understanding was that the GPS gave absolute accuracy. I really only need relative accuracy. Can I not just use relative GCPs and physically (tape measure and dumpy level) plot their coordinates relative to each other?

If you have access to something like ArcGIS or QGIS then it is possible but the accuracy is still questionable. In that case, you would georeference the image to an actual image (Google for example). However, you are more likely to induce more error by this method unless you are skilled (this method can further distort the imagery you collected). A control point is a known reference point on the surface used to tie specific points to a geographic location with a good to great degree of accuracy. For them to be effective and work with the processing software suites, you are going to need a long/lat (northing/easting) and elevation to create an accurate orthomosaic.

"Absolute accuracy" is a fallacy. Your accuracy depends on a lot of factors but the most significant one is the quality of your receiver. They come in various grades from the meter variety down to the millimeter variety. If you are using a prosumer drone off the shelf the unit accuracy can usually be measured in feet unless you have an RTK version which are usually centimeter grade. Atmospherics, multipathing, human error, EMI, bad luck and last night's activities all come into play to hose the solution as well. The only 'absolute' in GPS is that when you need it to work most, it might not.