Problem dealing with service area analysis

3627
19
09-11-2017 04:58 AM
G_shashiReddy
New Contributor II

Dan_Pattersonmitchh3001_jianhttps://community.esri.com/community/gis/analysis/network-analyst?sr=search&searchId=4adf3259-df94-4...

Can any one help me with this problem

I was performing Service area analysis, but I am facing this problem, the buffer is getting created along only one street. So plz help me solve this problem.

Adding to this I am also facing another problem saying

Warning: Location Facilities" is on a non-traversable network element position.@

0 Kudos
19 Replies
AnanMaalouf
New Contributor II

I'm having a similar problem.

I'm trying to create service areas of 200 meters around the bus stations (walkable distance).

While setting the break in the properties on 200 meters, and the trim = 100, I get much larger results than I would imagine (I thought it might arrive to maximum 300 meters distance). What might be the problem?

0 Kudos
MelindaMorang
Esri Regular Contributor

Polygon Trim Distance controls how close service area polygons should be from the reached network edges. The trim distance refers to the distance from the road the service area polygon should extend when no other reachable roads are nearby, similar to a line buffer size. It roughly models how far back from the road centerlines is considered reachable from the road. This is useful if the network is very sparse and you don't want the service area to cover large areas where there are no features.

So, the Service Area first calculates which roads are reachable within your 200 meter walk distance from your Facility.  After determining the reachable roads, it then generates polygons around those roads using a trim size of 100 additional meters.  So, theoretically, your polygon should be at most 300 meters from your input point, sort of.

In your screenshot, however, I see that your Facility is somewhat far from the road.  When you add this point as a Facility in the Service Area analysis, it must "locate" on the network, meaning that it effectively snaps to the closest position on the road network, and that position is used as the start position for the calculation.  I suspect the point is about 49 meters from the road, so when it snaps to the road, the polygon that gets created is still within the 300 meter limit from the input point's network location.  It looks correct to me.

Only somewhat related: For Service Areas this small meant to model pedestrians walking, you should probably use Detailed polygons.  They are more accurate.  You can change the polygon generation settings in the Service Area layer's properties in the polygon tab.

AnanMaalouf
New Contributor II

Thank you for your answer, it helped a lot! I understand it more clearly now.

Would you say that for the pedestrian walkability service area, some shapes might be 'distorted'? Since each station connected to a route with direction/sidewalk, and the pedestrian needs to cross in crossing lines, so the trims on the different sides might vary a lot?

0 Kudos
MelindaMorang
Esri Regular Contributor

Yes, the service area quality will only be as good as the underlying network data.  If your network doesn't have information about sidewalks, pedestrian paths, and crosswalks, then it won't have that level of detail reflected in the output.  Unless you have separate sidewalk features, it will assume the traveler is on the road centerline, and the service area polygon will be generated outward from the centerline in both directions.

AnanMaalouf
New Contributor II

Thanks again!

I do have sidewalk, and even though I added all (sidewalks, paths, crosswalks) to the network, I don't see their features in the 'UseThisStreets' layer of the network.

1) Is there a good online tutorial for pedestrian analysis?

2) Wouldn't it make sense that the trim would gradually decrease as the distance increases? So the trim shouldn't be a fixed value(?)

0 Kudos
MelindaMorang
Esri Regular Contributor

I don't know what the "UseThisStreets" layer is, so I don't know how to answer that question.

As far as tutorials for calculating the area within a short walk time/distance of your bus stops, there are two that may be helpful:

Older blog post

Newer Learn lesson

For the Service Area trim, I understand what you're saying, and it kind of makes sense in some ways.  However, trim is really just used to control the size of the service area buffer to indicate how far from the road you can reach.  It's useful for things like fire trucks because you can use it to model how far from the road your hoses reach.  So you can drive for 4 minutes from the fire station, and all the way at the very farthest road you can reach, you can still spray water 100 meters from the road where you park the truck (or something like that).

AnanMaalouf
New Contributor II

Thanks again!

'Streets_UseThisOne1' is the layer that is created when a Network is created.

The tutorial refers to ArcMap Pro, I guess my problem is with setting up and connecting the Network from a different type of layers (sidewalks, stairs, paths, streets, etc). It would be really helpful if there is a tutorial that addresses this while focusing on pedestrian accessibility.

But now looking at it, with understanding the 'trim' and 'impedance' it does make sense. How hard would be editing the code of the tool and adding a function: trim(i)=impedance-distance_from_facility+x(input_trim)

0 Kudos
MelindaMorang
Esri Regular Contributor

Ah, okay.  Based on the street layer name, I'm guessing you're using the Add GTFS to a Network Dataset toolbox to create your network.  You do NOT need this tool to simply calculate pedestrian walk times around transit stops. If your goal is to create a map showing the area within a short walk time or distance from the transit stops, then the Add GTFS to a Network Dataset toolbox is not the right tool for the job.  That tool creates a network you can use to calculate the area reachable by public transit and walking from a given location.  It includes both riding and walking, which, if I understand your workflow correctly, is not what you want.  If you just want to calculate walking areas around transit stops, you should use an ordinary network dataset that does not include the public transit stops and lines.  Does that make sense?

So, create a network that includes your streets, sidewalks, and crosswalks.  Then, add your transit stops as "Facilities" in the Service Area analysis layer.  Solve the Service Area.  Done!

AnanMaalouf
New Contributor II

That what I ended up with doing! Thanks..

0 Kudos
MelindaMorang
Esri Regular Contributor

Regarding trim, there is no way for a user to alter the way trim is calculated.  That's deep within the Service Area code.  I guess you could calculate Service Area lines and try to construct your own line buffers using some formula, although even that doesn't seem trivial.