TM1.6.0

Major changes since TM1.5

Work from home

Implement telecommuting (also known as working from home or WFH) into a simple "submodel" of the Coordinated Daily Activity Pattern model. For each worker (with a work location), that worker's choice to work from home is made via a linear function based on the log of the person's household income; the model specification is additionally segmented by home county and employment industry. These models were estimated using post-pandemic PUMS data.

Since the worker's employment industry is unknown, the worker's work-from-home probability is the weighted average of employment industry jobs at that worker's work location TAZ.

The work from home choice (wfh_choice) is stored as a person attribute, and written to the disaggregate person output file.

More information can be found here: Asana task: WFH model adjustment & validation - this is internal only.

Coordinated Daily Activity Pattern model calibration

Previously in TM1.5 the WFH workers were modelled within the CDAP universe. With the new "Simple WFH model" in TM1.6, WFH workers are taken out of the CDAP universe so we need more of the remaining workers to travel to work. This calibration ensured the number of workers going to work remain unchanged between TM1.5 and TM1.6. More information can be found in utilities/telecommute/calibrate_CDAP_WFH

Mode choice

Since the mode choice models were estimated/calibrated to pre-pandemic data, the original specification has embedded within it pre-pandemic mode choice behavior. As such, the following configuration has been added to represent post-pandemic changes to mode choice:

• Split SharingPrefFactor into Work vs NonWork and make this transit-hesitance specific (e.g. remove application to TNC mode choice utility). These are now called Work_Transit_Hesitance and NonWork_Transit_Hesitance
• Add BART_Transit_Hesitance (despite its name, this factor is being applied to BART, Caltrain and ferry in TM1.6)

California HSR Demand

Updated forecasts of access/egress trips to/from Bay Area CAHSR stations have been incorporated. Detailed documentation can be found in the high-speed-rail-external-demand/Readme and in the Asana task: Update interregional rail (CAHSR) travel assumptions - this is internal only.

Model calibration/validation

The focus of the model calibration/validation effort has been on accurately representing the change between 2015 and 2023

• Roadway valdiation using data from Caltran's Performance Measurement System (PeMS)
• Transit validation using data from the National Transit Database

Full Changelog: TM1.5.2.6...TM1.6.0

TM1.5.2.6

Changes since TM1.5.2.5:

New feature added specifically for NextGenFwy Study

• Implement means-based cordon discounts (for tolls and transit fares) by @lmz in #55
• Various new tools to support the Next Gen Freeways Study in /utilities/NextGenFwys

Repo organization (TAZ Data land use scripts and files are moved from the Petrale repository)

• Move files from petrale by @theocharides in #65
• Move files from petrale (Pt 2) by @theocharides in #66

This release is used in STIP2024.

Full Changelog: TM1.5.2.5...TM1.5.2.6

TM1.5.2.5

Changes since TM1.5.2.4:

• Update to operate within python3 conda environment called tm15-python310
• Minor updates to the Truck Model:
  • To improve truck validation, added 4 levers to move shares of IX/EX trips to trucks, IX_EX_TRK_[VSM,SML,MED,LRG]_SHARE, in IxTimeOfDay.job
  • To reduce what appears to be an oversensitivity of the truck trip distribution model to tolling scenarios, updated the truck trip distribution model to use a blended travel time for its gravity LOS term. The parameter, TRUCK_DISTRIB_LOS_TOLL_PART, is between 0.0 and 1.0, where 0.0 means to use the non-tolled (free) time only, and 1.0 means to use the tolled time only. We'll use 0.5 in this release.
• A lot of utilities to support the Next Gen Freeways Study
• Documentation fixes

TM1.5.2.4

Changes since TM1.5.2.3:

• Trip mode choice UEC fixes:
  • c_ivt_exp is reset to 1.00 * c_ivt
  • c_ivt_com is reset to 0.70 * c_ivt
  • Validation workbooks are available on Tableau public
    • Transit validation using legacy data
    • Transit validation using the consolidated transit passenger survey data
    • Roadways to PeMS
    • Roadways to Caltrans
• Validation scripts were updated to include the consolidated Transit Passenger Survey and APC data from AC Transit in prepare-validation-data
• Minor bug fix in PBA50 telecommute strategy script
• Operational enhancements and clean up for Plan Bay Area 2050:
  • changes committed between the release of TM1.5.2.3 on Dec 30, 2020 and Dec 9, 2021 (see details)
• Operational enhancements for TIP:
  • Added TIP to the project list in RunModel.bat
  • Updates to EMFAC prep scripts for air quality conformity applications in emfac_prep.py
• Initial files and scripts for NextGenFwys Study:
  • Model run log (ModelRuns.xlsx)
  • Metric scripts (Metrics.py and Metrics.twb)

This will be the version used for NextGenFwys.

TM1.5.2.3

Changes since TM1.5.2.2:

• Operational enhancements:
  • SetUpModel_PBA50.bat
  • check-setupmodel
  • Move bike constant for Complete Streets Network into config
  • EMFAC prep automation/streamlining
  • updated config for AOC and AV
• Strategy implementation for Final Blueprint:
  • Expand Transportation Demand Management Initiatives - parking fees input prep and parking summaries
  • Expand Commute Trip Reduction Programs at Major Employers - telecommute and preprocess

This is the version used for PBA50 Final Blueprint modeling.

TM1.5.2.2

This release includes updates that were used for the PBA50 Draft Blueprint model runs
Changes from TM1.5.2.1:

• AutoOwnership.xls includes config for 30% AV share target
• new script for setting up Draft Blueprint model runs - note this copies a few files from master that were not in TM1.5.2.1; these files have been committed into this release
• UseTollDist setting to trigger updateUECsToUseTollDist.py was removed since it was not necessary (all ODs had non-toll paths)
• ExtractKeyFiles.bat expanded to include more a few more files
• QAQC scripts added in postprocessing

TM1.5.2.1

Changes from TM1.5.2.0

• further mode choice calibration for updated 2015 auto operating cost

TM1.5.2.0

Changes from TM1.5.1:

• small adjustments to calibration constants for AutoOwnership, DestinationChoice, (Tour)ModeChoice and TripModeChoice in order to improve roadway validation for the Draft Blueprint
• improvements to calibration and validation related scripts, tools and process

TM1.5.1.11

Changes from TM1.5.1.10 (https://github.com/BayAreaMetro/travel-model-one/releases/tag/TM1.5.1.10)

• Don't force drive-alone tolls to be nonzero in AM, PM (384985b)
• IXEX update - handle the XX trips between 1461 and 1462 and update 2015 values slightly (7da8389)
• Add tollclass to avgload5period.csv and avgload5period_vehclasses.csv (7272cfd)

TM1.5.1.10

Changes since last release (TM1.5.1.9.1 ):

• Fix bug in core related to trip time of day (7abb373)