Service Network Area Polygons are longer than their distance limit

388
9
Jump to solution
12-10-2020 12:30 PM
VictorManuel
New Contributor II
Hello all!


I want to check the percentage of a city that is accessible to a forest (in green in the image) within a 2 km range from its entry points (purple dots). My first approach consists in creating buffers with a 2 km radio (blue circles) from the entry points. For a more realistic one, I am using Service Network Areas (ArcGIS 10.7) to use the network of streets of the city with the same distance limit (orange polygon). I got the network of streets of the city from openstreetmap and then I cleared it from rivers, streams, electric lines, railways and such. From the polyline shape I created a new network Dataset with only length in meters as attribute, and then created a New Service Area and loaded the 10 entry points of the forest as facilities. For the polygon creation I am merging the polygons and not trimming them (I'm also using Disk as overlap, even though I don't think this is important in my case, since I have only one break). The default break is 2000 m, which is the limit I defined for my study.

As you can see in the image, the maximum distance that could be covered should be the direct approach using the blue circles buffers. Nevertheless, the Network Service Area exceeds this maximum direct distance on some areas, reaching a distance of 2600 m from the entry points, for example. Whereas the limits of the Network Service Area polygon should be shorter than the direct approach, since the distance is being measured by using the network of streets. I am new in using this tool, therefore I don't know if I am overlooking something or making a mistake in my procedures.

I would greatly appreciate your help and comments.

Cheers,

0 Kudos
2 Solutions

Accepted Solutions
DavidPike
MVP Regular Contributor

There may be a projection issue, The length attribute of the roads may be mismatched with the dataframe projection if you've done a planar buffer. Try creating geodesic buffers in the first instance if you haven't already done so.

View solution in original post

DanPatterson
MVP Notable Contributor

Been around for some time, but there are little to no warnings about whether appropriate "fixes" are being used.  Just don't rely on webbie measures 😉

Measuring distances and areas when your map uses the Mercator projection (esri.com)

FAQ: Why are my map, distance and area measurements wrong when using WGS 1984 Web Mercator? (esri.co...


... sort of retired...

View solution in original post

0 Kudos
9 Replies
DanPatterson
MVP Notable Contributor

Is this a web mercator projection?


... sort of retired...
0 Kudos
VictorManuel
New Contributor II

Yes, Dan, it is. I will change the projection to see if the same mistake keeps coming.

0 Kudos
DanPatterson
MVP Notable Contributor

The reason I ask is Web Mercator is useless for distance and area measures.  Make sure you are buffering, running analyses and measuring in a projected coordinate system that is appropriate


... sort of retired...
VictorManuel
New Contributor II

Thank you, Dan. Indeed, it was the one that I got from the openstreetmap transformation and I kept it for simplicity reasons, I didn't know it was such a bad one. I will re-project and check again the results to come back to you.

Cheers,

0 Kudos
DanPatterson
MVP Notable Contributor

Been around for some time, but there are little to no warnings about whether appropriate "fixes" are being used.  Just don't rely on webbie measures 😉

Measuring distances and areas when your map uses the Mercator projection (esri.com)

FAQ: Why are my map, distance and area measurements wrong when using WGS 1984 Web Mercator? (esri.co...


... sort of retired...

View solution in original post

0 Kudos
VictorManuel
New Contributor II

Thank you, Dan! I converted everything to UTM33N (which applies for my case study) and now everything makes sense. See the pic attached. Rookie mistake haha

Thanks again!

UTM33N updatedUTM33N updated

DavidPike
MVP Regular Contributor

There may be a projection issue, The length attribute of the roads may be mismatched with the dataframe projection if you've done a planar buffer. Try creating geodesic buffers in the first instance if you haven't already done so.

View solution in original post

VictorManuel
New Contributor II

Thank you for your suggestions, David. I will check all my projections and look into geodesic buffers to come back to you.

0 Kudos
VictorManuel
New Contributor II

I have re-projected everything to UTM33N and now the Service Network Area makes sense. I also used geodesic buffers this time. Thank you for your help!

UTM33N updatedUTM33N updated

0 Kudos