Vertical Datum Transformations: CGVD13 to WGS84

345
4
Jump to solution
02-09-2021 07:39 PM
Labels (2)
Schillz
New Contributor II

Hello,

I'm wondering if it is necessary to apply a transformation from CGVD13 to WGS84? I have a large dataset of hydrographic soundings obtained with GCS WGS84 and a reference water surface elevation (WSE) obtained with CGVD13. I need to subtract the WSE from all depth soundings to get a true bed surface elevation. I can't find any references online that provide a simple or straightforward answer. 

From what I gather here, Height Reference System Modernization (nrcan.gc.ca), CGVD13 uses the CGG2013a realization which is 'synonymous' (?) with NAD83(CSRS)/ITRF2008 realizations. Then from other sources, WGS84 is 'synonymous' with ITRF2008. So does that make the CGVD13 and WGS84 ellipsoid near equal?

I'm relatively new to this topic and have found it quite a struggle to wrap my head around these seemingly endless transformations and coordinate systems...

Any help is greatly appreciated!

Cheers,

0 Kudos
1 Solution

Accepted Solutions
BojanŠavrič
Esri Contributor

Thank you, @Schillz, for your question and @Pål_Herman_Sund for your assistance. 

First, the theory...

As @Pål_Herman_Sund already said, "WGS 1984 and the CGVD 2013 are two very different things." CGVD 2013 (WKID::6647) is gravity-related vertical coordinate system (datum). WGS 1984 is geographic coordinate system (WKID::4326) and vertical coordinate system (WKID::115700) with positions and ellipsoidal heights, which you natively get from GPS observations. The first realization of CGVD 2013 is through Canadian Gravimetric Geoid model of 2013 – Version A (CGG2013a) in NAD83(CSRS) and ITRF2008. This basically means that geoid undulations (difference between ellipsoidal heights and orthometric heights) are provided between NAD 1983 (CSRS) or ITRF 2008 ellipsoidal heights and CGVD 2013 orthometric heights based on positions in NAD 1983 (CSRS) or ITRF 2008 geographic coordinates

To correctly transform between CGVD 2013 and WGS 1984 heights, it can be done in two steps:
GCS: NAD 1983 (CSRS) / VCS: CGVD 2013
<- - - vertical transformation with geoid model - - -> 
GCS: NAD 1983 (CSRS) / VCS: NAD 1983 (CSRS) 
<- - - 3D geographic transformation(s) - - -> 
GCS: WGS 1984 / VCS: WGS 1984


Note, your CGVD 2013 height data must be in NAD 1983 (CSRS) or ITRF 2008 positions, so you can transform them with the geoid model.

 

Now, the practice...

In ArcGIS Pro, you can transform your 3D vector and raster data with Project and Project Raster tools, by checking Vertical checkbox in the tool. The vertical checkbox becomes available when you add 3D data to the tool and set horizontal and vertical output coordinate systems.

BojanŠavrič_0-1613070348890.png

Unfortunately, we don't have any vertical transformations for Canada yet. Vertical transformation for CGVD 2013 is on our list, but some other projects came up and we haven't been able to get to it.

View solution in original post

4 Replies
Pål_Herman_Sund
Occasional Contributor

Hi,
to get the discussion going 🙂 many people in this forum have fare more knowledge in the field of geodesy than me. I have seen that being a bit "sluggish" in a reply often get the discussion going 🙂

WGS84 and the CGVD13 are two very different things. WGS84 is a mathematically defined earth-centered earth-fixed (ECEF) datum. The CGVD13 is a gravity-related datum/geoid model. The geoid model gives (at any point within its valid extent) the height difference between the CGVD13 and the reference ellipsoid

See https://www.nrcan.gc.ca/maps-tools-publications/maps/height-reference-system-modernization/9054#two

So is the "CGVD13 and WGS84 ellipsoid near equal?" No - there is a (geoid) separation between the two. I dont know the difference. For the same separation in Norway we have 20 - 50 meters (geoid above ellipsoid). 

cheers, pH 

Schillz
New Contributor II

Thanks pH,

Your description helps clarify that a transformation is indeed necessary. Which leads me to the next question of how do I apply a transformation between these two vertical datums? CGVD13 --> WGS84. Currently, my research has only stumbled upon a transformation from CGVD1928-->CGVD2013.

BojanŠavrič
Esri Contributor

Thank you, @Schillz, for your question and @Pål_Herman_Sund for your assistance. 

First, the theory...

As @Pål_Herman_Sund already said, "WGS 1984 and the CGVD 2013 are two very different things." CGVD 2013 (WKID::6647) is gravity-related vertical coordinate system (datum). WGS 1984 is geographic coordinate system (WKID::4326) and vertical coordinate system (WKID::115700) with positions and ellipsoidal heights, which you natively get from GPS observations. The first realization of CGVD 2013 is through Canadian Gravimetric Geoid model of 2013 – Version A (CGG2013a) in NAD83(CSRS) and ITRF2008. This basically means that geoid undulations (difference between ellipsoidal heights and orthometric heights) are provided between NAD 1983 (CSRS) or ITRF 2008 ellipsoidal heights and CGVD 2013 orthometric heights based on positions in NAD 1983 (CSRS) or ITRF 2008 geographic coordinates

To correctly transform between CGVD 2013 and WGS 1984 heights, it can be done in two steps:
GCS: NAD 1983 (CSRS) / VCS: CGVD 2013
<- - - vertical transformation with geoid model - - -> 
GCS: NAD 1983 (CSRS) / VCS: NAD 1983 (CSRS) 
<- - - 3D geographic transformation(s) - - -> 
GCS: WGS 1984 / VCS: WGS 1984


Note, your CGVD 2013 height data must be in NAD 1983 (CSRS) or ITRF 2008 positions, so you can transform them with the geoid model.

 

Now, the practice...

In ArcGIS Pro, you can transform your 3D vector and raster data with Project and Project Raster tools, by checking Vertical checkbox in the tool. The vertical checkbox becomes available when you add 3D data to the tool and set horizontal and vertical output coordinate systems.

BojanŠavrič_0-1613070348890.png

Unfortunately, we don't have any vertical transformations for Canada yet. Vertical transformation for CGVD 2013 is on our list, but some other projects came up and we haven't been able to get to it.

View solution in original post

Schillz
New Contributor II

Thanks @BojanŠavrič for the assistance and clear explanation. This is certainly helpful and answers my question concisely!