Calculating Sinuosity Properly

[EdwardGriffin]

Hi

I am creating a network dataset and one of the impedences to travel time I want to use is sinuosity or the bendyness of a road segment.

It's pretty important that this is included as roads in New Zealand go over mountainous terrain and road bendyness is a real impendence to travel time.

It is easy to calculate sinuosity by using Hawthes tools. However, sinuosity is determined by how many line segments make up a line. If only one segment makes up a line the sinuosity will always be one. This in effect means that sinuosity for some roads is understated as actually there are 5 line segments - they are just not linked together to make up a line.

What I want to create a dataset where all segments between two junctions are linked to form one line

Does anybody have any ideas how to do this, i'm struggling to think of anyway this will be possible?

Cheers

Ed

[AlexeyTereshenkov]

Hi Edward

I am a bit puzzled. Could you please clarify what exactly do you want to achieve?

As far as I got you, Dissolve Network tool might help (do you use ArcGIS 10?). But again, I might got you wrong.

[EdwardGriffin]

Hi Alex

What I want is to only have one road segment between each junction on my road dataset. At the moment sometimes I have up to five road segments between two junctions and this causes errors when I want to measure sinuosity because the segments are usually all straight (but make up a bendy road when put together)

Not sure if that makes any sense?

Cheers

ed

[AlexeyTereshenkov]

Hi Edward

I see. I assume that each road segment, i.e., a feature, has a specified speed value and other attributes. How are you going to deal with this? Are you fine with the idea of merging two adjacent road segments and assigning an average speed value and somehow a summed up travel time range?

And just one more thing. When you say junctions, do you mean junctions that are being created automatically when building a network dataset?

I will take a closer look at the data tomorrow at office.

[EdwardGriffin]

Hi Alex

Yes your right each segment does have specific attributes.

What i was going to do is calculate the overall sinuosity for each new segment between two junctions and then apply that back to the individual road segments

So the final dataset will still have all the orginal segments with their orginal attributes but the true sinuosity value also be included.

(if that makes sense).

In terms of junctions, yes the junctions that are created in NA but only the ones that have >2 roads converging at them - as i think the junction s just mimic the road segments.

Cheers

Ed

[JaySandhu]

Ed,

As Alex pointed out before, the GP Dissolve Network tool is what you want to use to remove the psuedo junctions (i.e. junctions with valency of 2). Then you can compute your sinuoisty property and apply it back to the original set of edges. Read more here:
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#/Dissolve_Network/00480000000w000000/

Jay Sandhu

[EdwardGriffin]

Hi Jay and Alex

I have only just noticed your replys. Thanks for your help on this - the dissolve function looks to be exactly what i'm after.

Cheers

Ed

[AlexeyTereshenkov]

Hi Edward,

Glad it played out well for you. If you already have got ArcGIS 10, you might take a look at those new GP tools in Network Analyst toolbox Jay and other people from Network Analyst team provide us with. I use them on a regular basis and they are great!

[EdwardGriffin]

Hi All

I have been trying to dissolve my road network but the I keep getting the same error on it: 030083 : Failed to dissolve the network dataset.

I have tried to trace this error and it is hard to understand what the problem is. I don't think there is anything wrong with the edges or the connectivity, and it doesn't have any elevation fields. It comes up short when it is trying to 'dissolve the feature chains'.

Any advice on this would be very gratefully received

cheers

Ed

[PaulBeere]

Yo' Ed

I solved your problem. Let me guess, you didn't assign a Z Coordinate System for the Feature Dataset that contained your road network layer? Was having the same problem with the Dissolve Network tool falling over at 'Dissolving feature chains', and I also got the corresponding '030083' error. Tried changing the Dissolve Network tool's environmental variables so that 'M Values' and 'Z Values' were set to 'Disabled', but that didn't work. I then tried adding a proxy Z Value column to my roads layer with all features assigned a value of 0, and setting this as the Z attribute. That just threw up an entirely new error message.

Finally, I tried assigning a Z Coordinate System for the Feature Dataset that contained my road network layer and now it is working. I did not change any of  the environmental variables. Turns out the GP tool was looking further up the chain for the Z geometry.

You are welcome

Paul