Shift in the local coordinates


I am not sure if this ever been addresses here in the community but my question is the following. I have done several tests over the known reference points and unfortunately the measurement over one of them has a considerable shift both in horizontal (~40cm) and vertical (~15cm) direction. The reason for this shift appeared to be an earthquake that happened in the area. A company that provides us with the RTK correction said that because of the earthquake, people who measure there are using some local calibration file in their Trimble devices. I am not sure how to approach this problem with Emlid because I am not a professional surveyor. The next measured known point in the 30 km distance does not have this shift so it looks like the shift is local and decreases the farther we go from the impacted area. The earthquake happened in 2008.

Anyone worked in the similar environment and have some suggestions to what can be possibly done in this case?


1 Like

Then don’t use that one?

The problem is that it applies to the whole town and the nearby areas. I just used that particular point to chech the measurement. We need to find the way to correct for these discrepancies. The guys that use trimble have some calibration file for the whole are but I have no idea how I can utilize this in my measurements with emlid.

I am just started to research this issue and there is a lot of things that I am learning along the way.

1 Like

Contact NGS, they will know what what to do if anything. I suspect this is operator error on the provider side.

I’ve never heard of this at all


Seems like a new permanent reference mark would of had to have been established even if in fact the mark was disturbed? If the others check out nearby, then you could temporarily set or compare using the data on record prior to it being disturbed?

I will meet with a local surveyor this week about this issue. But what he basically said is that the whole are has shifted during the earthquake and all the coordinates have some differences now. Everyone know about it and use their own base. Every major town has used their own local system. Great! Now I have to figure out what I can do with emlid to account for all this.

Hi @myka,

Emlid Flow also supports site calibration. If you have the known points, reobserve them with your Reach device and apply localization. Will this workflow work for you?

1 Like

Hi. Yeah I know of the site calibration option (which I haven’t tried yet) but unfortunatelly it will not suit our workflow. We are using Arcgis Field Maps and Emlid GNSS for high-precision survey and Field Maps does not really have any options to do anything except the point collection. All the settings are done in the Emlid flow (ntrip service and internet settings). Not sure if it even streams the localization data if I move to the Field Maps. I guess it only works in the Emlid Flow environment.

I spoke to a surveyor yesterday and they said that they use their own base with local grid. It works perfectly for them in the context of a local construction sites. But our assets in the GIS cover a large area and have the same CS so we need to find a way to somehow reflect those. Interesting indeed…

I am wondering if perhaps you should consider an outside in approach.

CORS Stations, such as the one providing the RTK Base (hopefully, they are in my part of the world), are basically all tied together on the ITRF between themselves with a very high confidence level of where they are in, in the world (which version is not so important for this discussion). The same is true if you were to send your own base data into AUSPOS or OPUS, you would get a result in the ITRF regardless of what event has occurred to your local situation.

If we then consider your area that was affected by an earthquake, it only affects a small portion of the world (in context), so you can then use these previously known points to work out how the earthquake has shifted the earth… which is of course what you have noticed, so how could you get around this?

If your survey covers a large area (City sized, so say at least 10km in any direction for example), you might be best to ignore the change due to the earthquake and just leave it is a simple RTK solution on your current State Datum and move on. Some time in the future, the Geoid for your State Datum will get updated, so too would the published co-ordinates of the marks that have shifted due to the Earthquake, but at least you have a solution for a point in time that the survey was taken.

In regards to others using their own localization, I would suspect this is due to the fact that in much smaller areas (such as at street level), the earthquake probably hasn’t had an effect of the ground distance between each point for say a subdivision or construction project, but in the global context, this area had shifted. So localization here would mean you can reobserve your control points and get a new relative position in the world.

Just remember, GPS Surveying is Global, so think big.

Hopefully this helps?


1 Like

GNSS surveying…

1 Like