LEC.314 Practical 3: Network Analysis (Part 1) Notes

Practical 3: Network Analysis (Part 1) Introduction

  • Network analysis in GIS uses network-based tools to model connections and flows through network datasets.
  • This session is split into 5 sections using different network datasets and analysis tools.

Learning Objectives:

  • Section 1: Route analysis for fastest routes on roads and canals using different network datasets.
  • Section 2: Find the closest facility (bus stop) to a chosen location.
  • Section 3: Determine journey length/time between sets of origins (pubs) and destinations (takeaways).
  • Section 4: Use service areas to identify:
    • Areas within an 8-minute drive of A&E in Lancaster.
    • Source areas upstream of a pollution incident in a river.
  • Section 5: Explore an ESRI location-allocation tutorial to find the best location for a new store in San Francisco, USA.

Learning Outcomes:

  • Explain why time of day can alter fastest travel routes in Lancaster.
  • Create a route that optimizes the order of deliveries for a traveling salesperson, apply a barrier to this route, and understand the changes this barrier can make to the route.
  • Compare and contrast routes based on 2 different network datasets and 2 different modes of travel.
  • Create a canal route with time impedance barriers.
  • Use the closest facilities analysis layer.
  • Use the Origin Destination Cost Matrix analysis layer.
  • Use the Service Areas analysis layer for travel distance from homes to A&E and pollution areas in rivers.
  • Access and complete an ESRI online tutorial on location-allocation.

Getting Started

  • Work with a pre-prepared map and associated data layers saved in a map package.
    1. Open ArcGIS Pro with a new Map, naming the project Network1 and saved in your LEC314 folder
    2. From Moodle right click on NetworkLancaster.mpkx and select Save Target As (or Save Link As) to save this into your ..LEC314\Network1 folder
    3. Import this map package to your new map in ArcGIS [Hint: Insert tab-> Import Map]
  • Ensure necessary layers are drawn and previous results are turned off.
  • Adjust zoom level for clear visualization.
  • Save Project before continuing.

1. Route Analysis

  • Calculates the quickest or shortest path along roads or other networks to visit stops (point locations).
  • Can be basic point-to-point routes or include barriers and incidents.
1.1 Roads
  • Using a network dataset from ArcGIS.com (available globally) to find optimum road routes.
  • Road network has time-based functionality using traffic data to model travel times in relation to specific dates or times (cloud-based, uses ArcGIS online credits).
  • A non-cloud-based network dataset produced from the Ordnance Survey’s Highways Network data set is also used. This accounts for most possible pedestrian routes and does not require any credits. A guide to using the Highways network data for your own project work is available on Moodle.
  • Steps for each network operation:
    • Create a new analysis layer.
    • Add relevant stops (and barriers if appropriate).
    • Run the analysis.
1.1 Fastest Route at different times of day

  • Explore how the route and associated travel time from the railway station to the university can vary at different times of day using the Route Analysis Layer.

    1. Click on the Analysis tab then on the dropdown for .
    2. Click Browse to change the Network Data Source so it is set to the cloud-based www.argis.com

  • The first network dataset that we will explore is an online resource provided by ESRI for use in ArcGIS Pro and ArcGIS Online.

    1. Click on the Network Analysis dropdown again and this time select Route
    2. Right click on the heading Route in the Contents pane and select Properties.
    3. Change the name to University to Railway Station.

  • Set parameters for the route:

    1. Click on the Route Layer tab at the top of the screen.
    2. Check that the Travel Mode is set to Driving Time.
    3. In the Data and Time group change the option from Not Using Time to Date & Time.
    4. Change the Sequence to Find Best.
    5. Set the time to be 22:00

  • Add stops to the route:

    1. On the Route Layer tab click Import Stops
      • For Input Locations select Railway Station then click OK [ignore any warning about the locations data being 3D and the network data source being 2D]
    2. Click Import Stops again
      • This time for Input Locations select University
      • Click OK
    3. Click Run Network Analysis on the Route Layer tab to solve the route

  • Total time and distance are reported at the bottom of the Directions pane.

  • Repeating the analysis with a different time (17:30) shows a different route is chosen.

1.2 Multiple Stops (travelling salesperson)
  • Concept expanded to consider multiple stops, optimizing the order in which they are visited (e.g., optimizing deliveries).
  • Example: van driver delivering pharmaceuticals products to six pharmacies in the Lancaster and Morecambe area. The start and end of their journey to be Junction 33 of the M6 motorway.
    1. Turn off (and collapse) the display of Railway Station to University
    2. Make a Route Analysis Layer and rename it Pharmacy Deliveries
    3. Repeat steps 9-13 but change Sequence to Preserve First & Last Stop and the travel time to 10:00
    4. Use Import Stops to create stops for J33, Pharmacies (field name as Name) and J33 again [adding this stop for a second time ensures it can be considered for the start and end of the route]
    5. Solve the route [Run Network Analysis]
1.3 Adding a Point Barrier
  • Network analysis used to find alternative routes around blockages (road works, accidents, flooding).
  • Implemented as point, line, or polygon barriers.
  • Example: accident on Greyhound Bridge (River Lune).
    1. Click on Import Point Barriers and select Input Locations to be Accident
      • Click OK
    2. Click Run to solve the route again
1.4 Driving vs Walking
  • Explore variation between driving and walking routes using the example of a route between a location on Graham Street and Rosebank Surgery.
  • Focus on shortest distance.
  • Time of day not accounted for, since journeys by foot do not vary significantly by time of day.
    1. Create a Route Analysis Layer renaming it Rosebank Driving
    2. Use the steps you have learnt to create and solve a Driving Distance route between the stops Graham Street and Rosebank Surgery.
    3. Solve this route again but for Walking Distance
    4. Check the basemap layer is set to OpenStreetMap
    5. Examine the map for a potentially shorter walking route – you will need to look very closely [red dashed lines show footpaths]
  • Limitations exist using the cloud-based network data for journeys by foot.
    1. Change the Network Data Source (see steps 4-5) to highwayslancaster.gbd\Network\Network_ND
    2. Create a Route Analysis Layer renaming it Rosebank Walking
    3. Use the steps you have learnt to create and solve a Walking distance route between the stops Graham Street and Rosebank Surgery.
    4. Turn off the display of Rosebank Walking
1.2 Canals
  • Network analysis can be applied to any set of connected linear features.
  • Example: calculate the time it takes to navigate from Lancaster to Glasson Dock by canal boat.
    1. Turn off (and collapse) the display of all layers
    2. Close any attribute tables that are open
    3. Change the basemap layer to GB Light Grey
    4. Turn on the display of Canal and zoom to the extent of this layer
    5. Change the Network Data Source (see steps 4-5) to Canal_ND
    6. Create a new Route Analysis Layer and rename the layer Canal Route
    7. Use your skills to solve the route for a canal boat between Lancaster and Glasson Dock Moorings. [Hint: Import Stops as Moorings]
  • Add a 15-minute time impedance for each of the locks along the canal.
    1. Use Import Point Barriers to account for the added cost (15 minutes) of Locks on the route. [Hint: Input Locations as Locks, then in the property list: BarrierType the default value should be Added Cost, and Attr_Minutes the default value should be 15]
  • Additional time can be allocated to barriers such as in this example, but also to stops, or connecting nodes in a network.

2. Closest Facility

  • The closest facility tool solves least-cost paths along a network.
  • Example: find the two closest bus stops with a service to the university, from a location of your choice within Lancaster and Morecambe using the Closest Facility Analysis Layer.
  • Use data from the Lancashire Local Council for bus routes and bus stops.
    1. Change the basemap layer back to OpenStreetMap
    2. Download shapefiles for the bus stop data and the bus route data, extract and save to an appropriate place on your H drive (e.g., your Network1 folder).
    3. Use Add Data to add the bus stop and the bus route data
    4. Turn on the display for the bus route data only
  • Spatially select the routes which pass through the university campus.
    1. Open the Attribute Table for the bus route data
    2. Zoom to Lancaster University campus
    3. Use Select tool [found on the map tab] to select the routes which pass through the university
    4. In the Attribute Table click to see only the selected values.
  • Select only those bus stops that have a service to the university using Select by Location, within a 20m distance of the route.
    1. Use Select by Location to select bus stops that are within 20m of the selected bus routes.
    2. Click on Explore to ensure you are happy with the stops selected.
    3. You can turn off the display of the bus route layer
  • Creating the Closest Facility layer.
    1. Change the Network Data Source (see steps 4-5) to highwayslancaster.gbd\Network\Network_ND
    2. From the Network Analysis dropdown select Closest Facility to create a Closest Facility Analysis Layer
    3. Rename the layer Nearest Bus stops
    4. Click on the Closest Facility Layer tab at the top of the screen, then change the Mode to Walk
    5. Increase the number for Facilities to 2 [this means the nearest 2 bus stops (facilities) will be found]
  • Select a start location (incident) from the code point layer (code points represent weighted centroids of postcode units).
    1. Turn on the display of Code Points
    2. Use Select by Attribute to select the postcode you wish to use (assuming you know the postcode. If you don’t know the postcode, zoom in to a location and use the Explore tool to click on the code point. This will show you a pop up containing details of the postcode)
  • Select a subset of bus stops from the current selection that are within 1 km of your selected postcode.
    1. We will assume that you wish to walk no more than 1km to a bus stop. Use Select by Location to select a subset of bus stops from the current selection that are within 1 km of your selected postcode. (This will reduce the number of facilities that ArcGIS has to load into memory and speed up processing considerably).
  • Find the route to the closest bus stops (facility) from your chosen location (incidents).
    1. Click on Nearest Bus Stops analysis layer in the Contents pane
    2. Click on the Closest Facility Layer tab at the top of the screen
    3. Use your skills to find the route to the closest bus stops (facility) from your chosen location (incidents). [Hint: Import Facilities as Bus stops and Import Incidents as Code Points, you can ignore the warning re Bus stops]
    4. Zoom to the extent of the newly generated route [Hint: within the sub-layers for Nearest Bus Stops right click on Routes and select Zoom to Layer].
    5. Return to the Map tab and click Turn off (and collapse) the display of Nearest Bus Stops

3. Origin Destination (OD) Cost Matrix

  • An OD Cost Matrix Analysis Layer calculates the journey length and time between sets of origins (max 1000) and destinations.
  • You can then rank destinations from each origin in order of minimum travel time or distance.
  • Example: determine which pub has the most takeaways within a ¼ mile walk.
    1. From the Network Analysis dropdown create an Origin-Destination Cost Matrix Analysis Layer and rename it OD Pubs-Takeaways
    2. Change the Mode for the OD Cost Matrix to Walk
    3. For Cutoff enter 0.25 [this will find all destinations within ¼ mile as the cost value for this dataset records distance in miles – note this is not always the case]
    4. Use your skills to calculate the journey length and time between pubs (origins) and takeaways (destinations). You will need to change the Field Name to BusinessNa in both the origin and destination layers to ensure they are included in the resulting pub-takeaway combinations.
  • Work out which pub has the greatest number of takeaways associated with it is to use the DestinationRank field.
    1. Under OD Pubs-Takeaways in the Contents pane right click on Lines and select Attribute Table
    2. Right click on the field heading DestinationRank and select Sort Descending
  • Turn off (and collapse) the display of OD Pubs-Takeaways and clear any selected features.
  • Use the toolbox ‘Solve Large OD Network Problem’ which runs on the ESRI cloud and can handle much larger datasets hence problems.

4. Service Areas

  • Creating a service area around a point, represents the maximum distance that can be travelled along a network.
4.1 A&E service area
  • Identify areas within an eight-minute drive of A&E at Lancaster Royal Infirmary.
    1. From the Network Analysis dropdown select Service Area to create a Service Area Analysis Layer and rename the layer A&E – 8 minutes
    2. Using your skills run the analysis to derive the area within an 8-minute (cut off) drive from the A&E facilities. [Hint: Drive Time, Away from facilities, cutoff as 8, Import facilities as A and E]. (If you wanted to compare multiple cutoff times you could set the cutoff to 8, 14, 20 for example and the analysis would generate three different polygons).
    3. Zoom to the extent of your service area.
  • Work out how many homes are within an 8-minute drive of the A&E department using code point data that represents postcode centroids.
    1. Use Select by Location to select Code Points that are completely within the service area you have just created. [Hint: the selecting features should be A & E - 8 minutes\Polygons]
    2. Open the attribute table for Code Points and note the number of records (postcodes) that have been selected.
    3. Right click on the field heading Dom_DP and click Statistics
    4. In the Chart Properties pane that appears look for the Sum under the Selection column (this is the total for the features you have just selected)
4.2 Pollution areas in rivers
  • Trace the river network upstream to search for potential source regions.
    1. Turn on the display of Rivers and zoom to the extent of this layer
    2. Change the Network Data Source to River_ND
    3. Create a Service Area Analysis Layer and rename the layer Rename the layer Pollution Area
    4. Use your skills to run the analysis to determine the 20km (20000 cut off) length of the river network that flows towards the pollution sampling facilities with an output geometry that has high precision. [Hint: Water, Toward facilities, cutoff as 20000, Import facilities as Pollution Source. Output Geometry group (Service Area tab) to High Precision]

5. Location- Allocation

  • The Location-Allocation analysis layer finds the best locations for facilities e.g. sports complex, to serve demand locations e.g. houses.
    1. Work through the ESRI tutorial on Location-Allocation which can be found here
Summary

During today’s session we have used a variety of ready prepared network datasets to demonstrate a selection of different types of analysis that can be conducted using network analysis.

Answers

  1. Yes
  2. It avoids busy sections of the A6
  3. 4 minutes
  4. 53km
  5. 5km (4mins) longer
  6. 788m
  7. Not really
  8. 398m shorter
  9. 146.3 mins
  10. 236 mins
  11. 175m or 0.21miles
  12. 20
  13. 25419