Reach RS3 / CH1903+ / LV95 + LN02 and SWIPOS

I’m looking for some advice setting up the Reach RS3 as a Rover
I am getting wrong elevation values in my local height datum.
This is my setup:
Connection via BLE to smartphone
NTRIP provider is SWIPOS, Mountpoint MSM_GISGEO_LV95LN02
Next NTRIP base station is only ~2 km away.
Solution status FIX (hence NTRIP corrections are coming in)
Pole height correctly set 1.8 m
Emlid Flow Project settings: CH1903+/LV95 + LN02 height

I then measure several known fix points from swiss national survey. The coordinates are well within expected accuracy (1-3 cm).
However the LN02 (Swiss elevation) value is way off (up to several dm).
When I measure the same point on different occasions/days, the original ellipsoidal height stays within a few cm.
I might add that our national survey office offers an online tool to convert WGS84 to our local height datum. This gets me the same values as Emlid Flow calculates. So I conclude that the Global ellipsoidal height is constantly offset, but why?

Here is a sample point I measured:

Fixpoint national survey (CH1903+/LV95 / LN02)
2781915.962 1185711.949 Height LN02 1562.35

9.82186585 46.79770662 1612.585 (Global)
2781915.961 1185711.944 Height LN02 1562.152

I measured additional points and can confirm that the offset is always around 0.2 m
I also double checked the pole height settings: 1.8 m in Emlid Flow → 1.8 m effective pole height. Antenna height is displayed as 1.934 m

I have probably found the error.
It seems that I have to use the
MSM_GISGEO_LV95LHN95 mountpoint instead of MSM_GISGEO_LV95LN02
This is quite counter-intuitive, since in the end I want to get LN02 heights. But I suspect that using the MSM_GISGEO_LV95LN02 mountpoint, I end up accounting twice for the difference (of about 20 cm in my region) between LHN95 and LN02. Once in the NTRIP stream and once in the rover.

I will do some more testing to be sure.
I think I might leave this thread for reference in case others have the same issue.

I talked to swisstopo an can confirm this.
Using MSM_GISGEO_LV95LHN95 does not shift the height in any way and is the “default” mountpoint. To use if you want to get LN02 height using CH1903+/LV95 + LN02 setting in the emlid app. This uses emlids internal chgeo04_HT_etrf geoid to directly get LN02.
Using MSM_GISGEO_LV95LN02 will already shift the incoming NTRIP correction data making your rover get its height already shifted. If you then use the above settings for emlid flow, your height gets shifted twice.

Hi Martin,

Sorry for the silence here, and thank you for sharing the updates about your survey! I’m glad to hear that it’s working fine with the other mount point.

What do you think, does the setting MSM_GISGEO_LV95LHN95 apply to all of Switzerland or is it region specific? I am currently using MSM_GISGEO_LV95LN02 in Wolhusen (LU) and some PP’s are centimeter accurate and some are off by 10-20 cm.

It should apply to entire switzerland.
In your region the difference between LN02 and LHN95 is only around 5cm so there must be something else.
What kind of control points did you use?

I only worked with one Rover, the NTRIP data was provided by the Lucerne base station. At the beginning of the survey, I focused on 3 known PPs and received these discrepancies.

Hi tinu80,
I just checked again. Looks like i found the source of the error. Previously, the measurement conditions were not ideal. Now I’m within measurement error.
So, THX!

Hi @szabuka78,

Thank you for sharing your experiences! I’m glad that everything’s working fine now. In case of any further questions, feel free to add a comment or create a new thread.