Sight lines in ArcGIS Pro

[MaximilianRehn]

Dear members,

a colleague and I try to analyze the visibility of wind turbines from several places in ArcGIS Pro.

In a first step we constructed “sight lines” by using the tool “Construct Sight Lines“between our observation points and the wind turbines (see figure 1).

Afterwards we tried to analyze the visibility of the wind turbines from our observation point by using the „Line Of Sight“ tool. We inserted the following datasets:

             - „in_surface“: a Digital Terrain Model

- „in_line_feature_class“: the mentioned sight lines between the observation point and the wind turbines

             - „in_feature“: a multipatch-feature of buildings

As you can see in figure 2 the output is just a single, uncut line, which crosses the multipatchbuildings.

We also tried to use as “in_surface” a Digital Surface Model and didn’t enter any “in_feature”. The result was a distorted line in a far too high position (see figure 3).

Our last idea was to cut our generated “sight lines” from figure 1 manually by deleting vertices or cutting lines on the points where they are touching the multipatch buildings. But the selected lines were moved upwards so that they are not touching the building anymore as you can see in figure 4.

Can someonehelp us with some proposals to solve our issues?

Thank you very much in advance!

[Robert_LeClair]

Maximilian - the Construct Sight Lines and Line of Sight output is expected behavior.  If you want to delete the lines that are classified Not Visible, you could do a Select by Attribute with the query VisCode=2 and delete those lines vs. manually deleting vertices or cutting lines.  This "should" resolve the issues of lines moving upward on your buildings.

[MaximilianRehn]

Thank you very much for your answer!

But we get only one line as output when we use the "Line of Sight" tool as you can see in figure 2. Actually it should be 21 lines because we have 21 wind turbines. The lines in figure 1 are the output of the "Construct Sight Lines" tool and manually classified in visible and not visible. Unfortunally this is a unprecise and time-consuming workflow.

[DanPatterson_Retired]

I don't suppose that forming a connecting polyline between the turbine locations has been tried.  It seems that you can get what you want if the locations were nodes of the polyline (densifying the output)

[Robert_LeClair]

Interesting.  My data is slightly different.  2 observers as point FC.  1 line FC as the target.  Output from the LOS tool is 389 lines vs a single line.  Could you try the Multipart to Singlepart to "explode" from a single feature to many features?

[MaximilianRehn]

Thank you very much for your answers, but maybe we are not talking about the same issue.

Principally we want to cut our sight lines between our obervation points and the wind turbines on obstacles like buildings. We want to get the same output as shown in the following image : http://desktop.arcgis.com/en/arcmap/10.3/tools/3d-analyst-toolbox/line-of-sight.htm. We inserted the datasets that are regarded in the tool description:

              - „in_surface“: a Digital Terrain Model ,

- „in_line_feature_class“: the constructed “sight lines” by using the tool “Construct Sight Lines“ between the observation point and the wind turbines,

             - „in_feature“: a multipatch-feature of Buildings.

But unfortunately the output is just a single, uncut line, which crosses the multipatch buildings. Our intended output were 21 lines because we have 21 wind turbines and therefore 21 sight relations. And they shall be cut on the buildings which form an obstacle.

Thank you in advance for your answers!

[Robert_LeClair]

Ooooohhhhhh...now I understand what you're talking about.  So the multipatch buildings are not "cutting" the line of sight lines created by the LOS GP tool nor creating a Point of Obstruction point feature class at the point of obstruction (i.e. the buildings).  I wonder if there's issues with your DTM, multipatch or something else.  Try this - ensure that all of your data is in the same projection vs. some layers projecting on the fly.  Second, convert your DTM to a new raster layer in the same projection as the other data.  Third, maybe there's an issue with your multipatch.  How was it created?