What projection is used to measure distance in Field Maps?

[RWATT]

Hello,

I'm wondering if anyone can share some insight into what's happening under the hood when linear distances are measured in field maps. 

I am using a web map that is built on a basemap using web mercator (WKID 3857).

I am measuring using the measure tool and using the automated length measurements of a hosted line feature layer in WKID 3857.

My location profile uses GCS NAD 1983 2011 (6318) as the GNSS Coordinate System and WGS 1984 Web Mercator Auxiliary Sphere (3857 as the Map Coordinate System. The datum transformation is ~WGS_1984_(ITRF08)_To_NAD_1983_2011. This is in Field Maps 23.1.0 Build 984.

When I measure a line on the map between two points collected with high-accuracy GPS, it is usually within an inch or two of the same distance measured with a tape measure. This obviously depends on the accuracy of the GPS, but it's close enough that I'm confident it is not measuring in web mercator. The distortion on this line is more than 30' when measured under web mercator in ArcPro.

This is good, but I want to understand how it's getting there. I had assumed it would be measuring in web mercator, because the base map and feature layer are in web mercator, and the location profile is using web mercator as its coordinate system.

Is field maps measuring geodesic distance? Using the GNSS Coordinate System? Using an on-the-fly projection based on the location profile extent, or map extent? Choosing an appropriate state plane? 

The closest related post I could find was this one, which doesn't really dive into projections:

https://community.esri.com/t5/arcgis-field-maps-questions/field-map-measurements-which-is-correct/m-...

Thank you,

[BenSundy]

Here is an Esri article kind of related to your question: Measuring distances and areas when your map uses the Mercator projection 

However, I would also like to better understand what's happening under the hood when using high accuracy GNSS and app with WGS84. I use a template GBD in a specific SRS for a data collection project, create a feature services from it for use in Field Maps which automatically gets reprojected to WGS84 when published (?), and then download and re project back into my desired SRS to perform analysis and provide to other project staff. When working with High Accuracy GNSS, a 3 ft shift could completely render your collected data inaccurate, so I wish the whole process was a little more clear.