Select to view content in your preferred language

Why are areas calculated in ArcGIS Pro different to the same calculation in ArcMap?

2486
6
Jump to solution
02-19-2023 05:06 PM
Labels (2)
LindsayRaabe_FPCWA
Regular Contributor II

I have 2 datasets that were put through a union and using the output I have then calculated the geometry to get a Hectare area figure. It's a broad layer covering a vast area of Western Australia. 

LindsayRaabe_FPCWA_2-1676854677989.png

 

In ArcMap, we calculated the geometry using 3 different coordinate systems and got 3 different areas (as we expected we would). 

LindsayRaabe_FPCWA_0-1676854533107.png

LindsayRaabe_FPCWA_5-1676855000128.png

 

LindsayRaabe_FPCWA_1-1676854554646.png

I have then repeated the process in ArcGIS Pro (same data sources and process) and got the below results - and not only do the figures differ from the ArcMap results, but they are the same from projection to projection (this is what originally got us looking closer as we expected different numbers). 

LindsayRaabe_FPCWA_4-1676854911175.pngLindsayRaabe_FPCWA_3-1676854686920.png

The Albers figure is closest to matching between the ArcMap and ArcGIS Pro figures which makes sense as it's a broader scale projection compared to Zones 50 & 51, but still differs. 

Can anyone explain either what we're doing wrong or why the results in ArcGIS Pro are seemingly consistent when they shouldn't be?

 

Lindsay Raabe
GIS Officer
Forest Products Commission WA
1 Solution

Accepted Solutions
sjones_esriau
Esri Contributor

Hey Lindsay,

"Or will geodesic fix this too?" 

Yes, Kind of. 

sjones_esriau_0-1676883352041.png

 

In the example above I've calculated the area in both Planar and Geodesic.

When calculating the planar area of the data in Zone 51, the geometry is projected from the globe to a flat extended plane which is centered on the central meridian of zone 51, somewhere to the east. 

 

sjones_esriau_2-1676884090536.png

This is why the data gets distorted the further you are away from the central meridian of the zone and why calculated planar area for Zone 51 is larger than Zone 50 (roughly where the source data is).

For the geodesic calculations, the geometry is still being projected to the zones, but now the extended zone wraps around the globe rather than being a flat plane. Since the GDA zones are uniform, the area and length calculations of features will be the same for geodesic measurements between zones.

This doesn't negate the need for zones however as Pro map views also operate in planar space unless specific tools or operations provide the options to create or calculate data differently. (e.g. create geodesic in editing, geodesic viewsheds, calc geometry in geoprocessing etc.). 

The mind bending thing is, like projections, all these area calculations are mathematically correct. You just need to use the one that is fit for purpose and record it when reporting. For large area calculations within Australia you can also try the GDA 94 Geoscience Australia Lambert projection or Albers.

(also upgrade Pro, if you are not seeing the dropdown options)

View solution in original post

6 Replies
DanPatterson
MVP Esteemed Contributor

geodesic calculations are good... I don't think ArcMap had them (last time I used it was some time ago)

Calculate Geometry Attributes (Data Management)—ArcGIS Pro | Documentation

If you want to check the Albers with arcmap, try a planar calculation method to see if they are similar.


... sort of retired...
0 Kudos
sjones_esriau
Esri Contributor

All the Area and Length calculations in ArcMap are planimetric.

If you redo the calcs in Pro using the 'Area' property in the dropdown, you should get the same result as ArcMap.

0 Kudos
LindsayRaabe_FPCWA
Regular Contributor II

Thanks for info @DanPatterson and @sjones_esriau . We didn't have the plan 'Area' option in our drop downs, so couldn't compare the planimetric and geodesic calcs, but I think I'm satisfied with the differing results. The only doubt left in my mind is around the difference between Zones 50 and 51. Wouldn't this still cause a difference in errors as the features when calculated in a zone that they don't fall within are further from the intended use area, and thus distorted? Or will geodesic fix this too? (in which case, are zones rendered pointless going forwards?)

 

R.png

Lindsay Raabe
GIS Officer
Forest Products Commission WA
0 Kudos
sjones_esriau
Esri Contributor

Hey Lindsay,

"Or will geodesic fix this too?" 

Yes, Kind of. 

sjones_esriau_0-1676883352041.png

 

In the example above I've calculated the area in both Planar and Geodesic.

When calculating the planar area of the data in Zone 51, the geometry is projected from the globe to a flat extended plane which is centered on the central meridian of zone 51, somewhere to the east. 

 

sjones_esriau_2-1676884090536.png

This is why the data gets distorted the further you are away from the central meridian of the zone and why calculated planar area for Zone 51 is larger than Zone 50 (roughly where the source data is).

For the geodesic calculations, the geometry is still being projected to the zones, but now the extended zone wraps around the globe rather than being a flat plane. Since the GDA zones are uniform, the area and length calculations of features will be the same for geodesic measurements between zones.

This doesn't negate the need for zones however as Pro map views also operate in planar space unless specific tools or operations provide the options to create or calculate data differently. (e.g. create geodesic in editing, geodesic viewsheds, calc geometry in geoprocessing etc.). 

The mind bending thing is, like projections, all these area calculations are mathematically correct. You just need to use the one that is fit for purpose and record it when reporting. For large area calculations within Australia you can also try the GDA 94 Geoscience Australia Lambert projection or Albers.

(also upgrade Pro, if you are not seeing the dropdown options)

LindsayRaabe_FPCWA
Regular Contributor II

Thanks for the detailed info. I think this answers our concerns adequately. FYI - my colleague was running 2.9.5 and I'm on 3.0.0. Waiting for 3.1 to rollout so we can all upgrade to that. 

Lindsay Raabe
GIS Officer
Forest Products Commission WA
0 Kudos
DanPatterson
MVP Esteemed Contributor

Can't find the link, but if memory serves, planar calculations treat true curves in the geometry differently than geodesic, so that may be a partial cause.  Latitude does as well. Also, the geometry coverage within the zone and where geometry is within the zone will have an influence.  Planar measures particularly for projections with a central meridian scale factor closer to 1 (like 0.9999) will approximate geodesic calculations better.  And another also, you are calculating for large areas, so geodesic would be better.


... sort of retired...