Linear Referencing

Problem

Linear data is typically thought of as a line between two points or nodes. That line may have a multitude of features along that segment between the two nodes that change many times. Linear referencing allows those points to be shifted depending on the attribute which is being analyzed. In this example exercise my client would like to determine if there is any relationship between two attributes (accidents and pavement quality) of a road in Pitt County, North Carolina.

Analysis Procedures

Methods

I started by creating a new customized ribbon for linear referencing adding 3 sections: Routes, Events & Add-In Tools. In the Route section I added the following tools: create routes, calibrate routes, and locate features along routes. In the Event section I added the following tools: dissolve route events, make route event layer, overlay route events, and transform route events. In the Add-Ins section I first had to download Linear Referencing Add-In Tool from the Texas DOT via GitHub and then I was able to add set from/to measures and the identify route locations tools.

Next I added the city and county boundaries and a feature class called some_routes. I used the Identify Route Locations tool to locate a specific route of interest (3000030). I then added the accident information table and then created a route event layer. The same process was applied to the pavement condition table. To determine if there is a correlation between the number of accidents and the pavement conditions on route 30000030 I used the overlay route events tool and determined that on this particular section of the road that there were more accidents on the sections that have a quality rating of 75 or higher.

Results

Application & Reflection

In this assignment, I learned how to analyze linear data by its various attributes. This type of analysis could be done to establish correlations between any number of linear attributes but in particular I can see where you could correlate water quality readings along a river with fish population survey data. You would need a table of the water quality readings and a table of the fish population samplings (from NC DEQ). You would create an event layer from each and then overlay the resulting route events to see if the water quality readings can account for any variations in the fish populations along the length of the river.