1 of 33

Forecasting Households & Jobs along the Wasatch Front using Urbansim

Josh Reynolds

Modeler

Wasatch Front Regional Council

2 of 33

Background

Education

  • B.S. Geographic Information Systems (University of Wyoming)
  • M.S. Geography (University of Utah)
    • “Comparing Urban Vegetation Cover with Summer Land Surface Temperature in the Salt Lake Valley”

​

Work History

  • GIS Analyst (DIGIT Lab)
  • Geospatial Programmer (Redcastle Resources)
    • Vegetation mapping
    • Post-Wildfire Burn Severity Analysis
  • Modeler (WFRC)
    • Land Use Modeling
    • Bike Demand Modeling
    • Equity-related mapping and support

​

3 of 33

Quick WFRC Intro

  • Metropolitan Planning Organization and AOG�
  • Regional Government
    • Facilitate Regional Visioning & Planning
    • Short-Range Transportation Programming
        • With UDOT & UTA
        • Next 4-6 years ($3-4B)
    • Long-Range Regional Transportation Plan
        • Through 2050 ($23B)
    • Economic Development Coordination�
  • Your Mayors, Commissioners, & Councilpersons�Are our immediate stakeholders (and Board)

as MPO

as AOG

4 of 33

~55% of Utahns

~20% of Utahns

Mountainland MPO Area

Pop. 645,000

(2019 est.)

Wasatch Front Regional Council (WFRC)

Pop. 1,867,000

(2021 est.)

55% of Utahns

Mountainland (MAG)

Pop. 673,000

(2021 est.)

20% of Utahns

5 of 33

414

Metropolitan Planning Organizations, by state

  • Mission: Ensure regional coordination and public participation in transportation planning and related decision-making processes
    • Policy Committee, Technical Committee,
    • Planning and Coordination Staff

6 of 33

Wasatch Front Regional Council (WFRC)

WFRC

Council:

Local Elected Officials,

Utah Senate & House,

UDOT, UTA, ULCT,

Envision Utah

Staff:

Planners,

Technical Staff,

Govt Relations,

Communications,

Finance

Policy, Decision-making, Funding Approval

Short & Long-Range Planning,

Transportation & Land Use Modeling, GIS

7 of 33

RTP: 20+ year planning horizon

UDOT/UTA/Local Govt

​

TIP: 4-6 year funded projects

UDOT/UTA/Local Govt

STP, CMAQ, & TAP project funding

UDOT/Local Govt

​

CEDS: Regional Framework

(US Commerce Dept/Local Govt)

8 of 33

Wasatch Choice: Collaborative Visioning & Planning

The Wasatch Choice Vision is our communities’ shared plan for transportation investments, development patterns, and economic opportunities

​

Key Strategies:

  • Provide Transportation Choices
  • Support Housing Options
  • Preserve Open Space
  • Link Economic Development with Transportation and Housing Decisions

​

9 of 33

Wasatch Choice Integrated Map

Transportation, Land Use, Economic Development & Quality of Life are all inter-related.

�Our interactive map should represent that integration….

wfrc.org/wasatch-choice-map

10 of 33

WFRC Models

11 of 33

WFRC Models

2 main models:

  • Wasatch Front Travel Demand Model (TDM)
  • Real Estate Market Model (REMM)

Simulate the future and allow for scenario testing

12 of 33

Travel Demand Model

4-Step Trip-Based Model

Models trips (by-purpose) on a zone-to-zone basis for all Traffic Analysis Zones in the modeling region

Models:

  • Roadway volumes,
  • Travel speed indicators
  • Transit route boardings
  • Vehicle miles traveled
  • Vehicle hours traveled
  • Transit/auto/non-motorized mode shares
  • Trip length costs

The results are used for:

  • Forecasting traffic volumes and transit ridership
  • Air quality and environmental impact studies
  • Prioritization of projects for funding

wfrc.org/traffic-volume-map/

13 of 33

Urbansim

An open-source urban simulation platform designed by Paul Waddell (University of California, Berkeley) and other collaborators

Used for modeling metropolitan land use, transportation, and environmental planning

​

14 of 33

Real Estate Market Model

  • The Real Estate Market Model (REMM) is our adaptation of the Urbansim framework, customized to the Wasatch Front region data and needs
  • Estimates and Forecasts Socioeconomic variables for the years 2019-2050 for the Travel Demand model
  • Written in Python
  • Hosted on Github

15 of 33

WFRC/MAG Models to Project Transportation Future

Data

current conditions future plans, and

county-level projections

Road & Transit Facilities

​

• Road Network

• Capacity & Lanes

• Routes & Schedules

• Future Projects

​

Housing & Demographics

• Population Characteristics

• Housing Mix & Distribution

• ‘Pipeline’ Housing Dev

• County Growth Projections

Real Estate Market

• Existing Land Uses

• Property Values

• Building & Lot Sizes

​

Local Land�Use Policy

Local Land Use Policy

• General Plans

• Masterplans/Visions

• Re-/Developable Lands

​

System Usage

• Observed Speeds

• Trip Origins/Destinations

•Transit Ridership

• Travel Surveys

​

Employers

Employment Locations

• Employer Location

• Employee Counts

• ‘Pipeline’ Commercial Dev

• County Growth Projections

16 of 33

WFRC/MAG Models to Project Transportation Future

Data

Models

current conditions future plans, and

county-level projections

regional simulations, using local information,

model the future through 2050

Road & Transit Facilities

​

• Road Network

• Capacity & Lanes

• Routes & Schedules

• Future Projects

​

Housing & Demographics

• Population Characteristics

• Housing Mix & Distribution

• ‘Pipeline’ Housing Dev

• County Growth Projections

Real Estate Market

• Existing Land Uses

• Property Values

• Building & Lot Sizes

​

Travel Demand Model

Simulates trips:

• To where?� • What mode?

• What route?

​

​

Land Use Model (WFRC/MAG)

Simulates development:

• Market Conditions

• Profitability� • Where, When, Intensity

​

​

Local Land�Use Policy

Local Land Use Policy

• General Plans

• Masterplans/Visions

• Re-/Developable Lands

​

System Usage

• Observed Speeds

• Trip Origins/Destinations

•Transit Ridership

• Travel Surveys

​

Employers

Employment Locations

• Employer Location

• Employee Counts

• ‘Pipeline’ Commercial Dev

• County Growth Projections

feedback

17 of 33

WFRC/MAG Models to Project Transportation Future

Data

Models

Forecasts

current conditions future plans, and

county-level projections

regional simulations, using local information,

model the future through 2050

household and job distribution; transportation system performance;

scenario evaluation

Road & Transit Facilities

​

• Road Network

• Capacity & Lanes

• Routes & Schedules

• Future Projects

​

Housing & Demographics

• Population Characteristics

• Housing Mix & Distribution

• ‘Pipeline’ Housing Dev

• County Growth Projections

Real Estate Market

• Existing Land Uses

• Property Values

• Building & Lot Sizes

​

Travel Demand Model

Simulates trips:

• To where?� • What mode?

• What route?

​

​

Land Use Model (WFRC/MAG)

Simulates development:

• Market Conditions

• Profitability� • Where, When, Intensity

​

​

Local Land�Use Policy

Local Land Use Policy

• General Plans

• Masterplans/Visions

• Re-/Developable Lands

​

System Usage

• Observed Speeds

• Trip Origins/Destinations

•Transit Ridership

• Travel Surveys

​

Employers

Employment Locations

• Employer Location

• Employee Counts

• ‘Pipeline’ Commercial Dev

• County Growth Projections

future:

feedback

Traffic Volumes & Speeds

City & TAZ Growth

Urban Form

Air Quality

18 of 33

Traffic Volumes & Speeds

WFRC/MAG Models to Project Transportation Future

Data

Models

Forecasts

Partners

• Planning Depts

• Tax Assessors

City & County�

• MPOs, AOGs

• Transit Agencies

Regional�

• UDOT

• University of Utah (GPI)�• Utah DEQ

• Utah Population Committee

• Workforce Services

​

State�

• Census Bureau

• FHWA & USDOT

• FTA

Federal�

• Consultants�• Data Providers

• Real Estate Experts

Private�

current conditions future plans, and

county-level projections

regional simulations, using local information,

model the future through 2050

household and job distribution; transportation system performance;

scenario evaluation

Road & Transit Facilities

​

• Road Network

• Capacity & Lanes

• Routes & Schedules

• Future Projects

​

Housing & Demographics

• Population Characteristics

• Housing Mix & Distribution

• ‘Pipeline’ Housing Dev

• County Growth Projections

Real Estate Market

• Existing Land Uses

• Property Values

• Building & Lot Sizes

​

Travel Demand Model

Simulates trips:

• To where?� • What mode?

• What route?

​

​

Land Use Model (WFRC/MAG)

Simulates development:

• Market Conditions

• Profitability� • Where, When, Intensity

​

​

Local Land�Use Policy

Local Land Use Policy

• General Plans

• Masterplans/Visions

• Re-/Developable Lands

​

System Usage

• Observed Speeds

• Trip Origins/Destinations

•Transit Ridership

• Travel Surveys

​

Employers

Employment Locations

• Employer Location

• Employee Counts

• ‘Pipeline’ Commercial Dev

• County Growth Projections

future:

City & TAZ Growth

Urban Form

Air Quality

feedback

19 of 33

Land Use Model Inputs

​

20 of 33

GPI County Control Totals

​

The Kem C. Gardner Policy Institute (University of Utah) provided us with county-level growth estimates and forecasts for each year between 2015-2060:

  • Population
  • Household
  • Job (by sector)

Our model attempts distributes these “control” values to the modeling space

  • Where will the population growth occur?
  • Where will office jobs be located?
  • What city has the most potential for growth?

​

21 of 33

Parcels and Buildings

Source: County Assessors

​

Parcels (modeling unit)

  • Acreage
  • Land value
  • Accessibility to households and jobs
  • Distance to highways and transit

​

Buildings

  • Square footage
  • Building type
  • Value
  • residential units
  • job spaces

Land Value

Building Type

22 of 33

General Plans

Source: City & County Plans, Master Plans

​

Zoning - what types of buildings are allowed to develop?

  • Single family, multifamily, industrial, retail, office, government, mixed-use, other

​

Capacity - How dense can we build?

  • Dwelling Units per Acre (Residential)
  • Floor-to-Area Ratio (Commercial)

​

Maximum DUA

Single Family Allowed

23 of 33

Households & Jobs

Households (the people/agents)

  • Synthesized using 2020 Census and Public Use Microdata Areas (PUMA) data
  • Income
  • Household size
  • # of cars
  • Workers
  • Children

​

Jobs (attractions)

  • Processed from Department of Workforce Services data
  • Only about 70% of jobs are placeable, the rest must be synthesized
  • Sector
    • Accommodation & Food
    • Healthcare
    • Government & Education
    • Manufacturing
    • Office
    • etc…

24 of 33

Local Feedback & Expertise

We incorporate feedback from city planners, developers, local officials, and other experts to improve our results

​

We also monitor websites like constructionmonitor.com and buildingsaltlake.com for new developments that we assert into our forecast

25 of 33

Land Use Sub-Models

​

26 of 33

Sub-Models in REMM

2019

​

​

​

​

​

​

​

​

​

​

​

2050

27 of 33

OLS (Ordinary Least Square Regression)

    • One model for each county
    • Estimated from observed data
    • Residential price
    • Commercial price

​

Examples of Relevant Variables

    • Size and age of structure, vacancy rate of the area, accessibility, availability of transit service

THE MODEL | LOCATION CHOICE MODELS

Price Model

27

| SLIDE

28 of 33

Pro Forma

    • Determines what can feasibly develop by estimating potential profit of each parcel
    • Also helps determine when a parcel is ready for redevelopment

​

​

​

​

​

​

THE MODEL | DEVELOPER MODEL

Real Estate Development Modules

28

| SLIDE

29 of 33

Multinomial Logit Model

​

Households:

  • Used to allocate new or displaced households to residences
  • One model for each income quartile and county

​

Employment:

  • Used to allocate new or displaced jobs to buildings
  • One model for each job sector and county

​

Examples of Relevant Variables

    • Age of structure, income of surrounding residents, traffic volume of adjacent roads, distance from freeways

THE MODEL | LOCATION CHOICE MODELS

Location Choice Models

29

| SLIDE

30 of 33

Outputs

One model run takes about 6 hours

A database of simulated parcels and buildings for each year

TAZ-level summary of socioeconomic variables

  • Number of Households,
  • Average HH size,
  • Total Population,
  • Average Income
  • Jobs by sector
  • Grade School enrollment

The average of 6 model runs is fed to the Travel Demand Model as an input

​

​

31 of 33

A few takeaways…

Data collection is challenging; knowing how to process tables with code can speed things up and help you stay organized

​

Forecasting is never perfect. Cities change their plans all the time, unexpected situations happen (e.g,. COVID-19). However, it is still advantageous to have some kind of plan.

​

Collaboration with other agencies and peers is essential for completing large projects

  • Data
  • Feedback

32 of 33

Questions?

Home Page:

wfrc.org

​

Map Gallery:

maps.wfrc.org

​

Data Portal:

data.wfrc.org

​

Github:

https://github.com/WFRCAnalytics

33 of 33

Resources

​

​