Integrating Aviation and Passenger Rail Planning (2015)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

196 ACRP 03-23 Air/Rail Diversion Model: User’s Guide a p p e n d i x

Outline This user’s guide for the ACRP 03 23 Air/Rail Diversion Model providers instruc
ons on how to install and use the model to evaluate scenarios. The input files are described in detail in the input files specifica
on sec
on, while the final sec
on of the user’s guide provides an overview of the code structure of the model for advanced users who might wish to edit the model. • Introducon • Model files and Installaon • Creang Scenarios • Eding Inputs • Running the Model • Viewing Output • Input File Specificaons • Code Guide

Introducon The ACRP 03-23 Air/Rail Diversion Model is a sketch planning model to provide model users with a quick response tool capable of evaluang a range of policy intervenons that affect choice of air or rail for long distance travel in a mega region sized corridor. The model is intended to provide a framework for managing scenarios, eding inputs, running the model, and viewing outputs. All inputs and outputs are in .csv format to also allow viewing and analysis in so‚ware other than the ACRP 03-23 Air/Rail Diversion Model. This users’ guide does not describe the model structure; this is covered in the technical appendix to the project report found on the enclosed CD.

Model Files and Installaon

Installaon • Requirements • The ACRP 03-23 Air/Rail Diversion Model is designed to work on a PC computer that has a recent version of Microso Excel installed (2007 or later). • Step 1: Save ZIP folder to Computer • The ACRP323_AirRailDiversionModel.zip folder includes all of the necessary files to run the model. To install the model, click on the Air/Rail Diversion Model menu buˆon. Save the zip folder labeled ACRP323_AirRailDiversionModel.zip to the appropriate locaon by clicking “ok.”

Installaon (Contd.) • Step 2: Unzip the ZIP folder • Unzip the ACRP323_AirRailDiversionModel.zip using a zip ulity installed on your computer (e.g., WinZIP). The zip file’s contents are inside a directory call “ACRP323_AirRailDiversionModel” • That directory contains: • R model scripts and R applicaon in the “_RModel” folder • File structure with model inputs and outputs, by scenario, in the “East Coast” and “West Coast” folders • Two spreadsheets: Excel spreadsheet graphical user interface (GUI) for current versions of Excel, and also Excel 2007

Open the Excel GUI • Step 3: Open the Excel GUI There are two versions of the excel GUI, one for Excel 2007 and one for more recent versions of Excel. Open the one that conforms to the version of Excel on your computer.

Test the installaon Step 4: Test the Installaon Test the installaon and connecon between the Excel GUI and R (open source soware that runs the model’s simulaon and is included with the model zip file). On the Main Menu sheet in the excel GUI, click on the Run Model button to launch a run. A command window will launch and the model’s progress statements will start to appear. Click the “RUN MODEL” button Command window with progress statements

Model Files: Models and Scenarios The model comes with two pre-developed models, for the East Coast and West Coast regions, which are contained in the “East Coast” and “West Coast” folders, respec
vely. Within each of those two folders are a list of scenarios. Ini
ally, there is only a “Base” scenario folder, but as scenarios are created (described in the following pages), addi
onal scenario folders will appear here. The screenshot below shows the contents of the “East Coast” folder – just the “Base” scenario folder at the moment.

Model Files: Inputs and Outputs Each scenario folder, for example, with in the East Coast/Base folder as shown here, contains folders containing .csv files for several different categories of inputs, and the “Base” and “Future” folders which contain outputs (once the scenario has been run). The ScenarioParameters.csv file contains several top level policy variables (described later in the user’s guide)

Model Files: R Model, Applicaon The _RModel folder contains the R folder, which is a version of the R applicaon that is used to run the model. The files in that folder are as follows: • ACRPData_.Rdata – R binary folder containing several input tabulaons • ACRPModel_.Rdata – R binary folder containing model coefficients • Log.txt – log file from a model run (added when the model is run in installaon) • ModelApplicaon.R – R script containing the model code • RunModel.R – R script to run the model (added when the model is run in installaon) • RunSpecificaon.txt – text file written by the Excel GUI providing variables to R (added when the model is run in installaon)

The Model’s GUI

GUI Workflow As noted above, the ACRP 03-23 Air/Rail Diversion Model is a sketch planning model to provide model users with a quick response tool capable of evalua€ng a range of policy interven€ons that affect choice of air or rail for long distance travel in a mega region sized corridor. Model users do this by running different scenarios. A typical workflow is as follows: 1. Run the base scenario (This represents the current situa€on – the business as usual case or reference case) 2. Create an alterna€ve scenario 3. Edit the alterna€ve scenario’s inputs to represent the policy to be tested 4. Run the alterna€ve scenario 5. View the alterna€ve scenario’s results 6. Repeat steps 2 5 for all of the policies to be tested 7. Compare the results of the policy runs

Elements of the GUI The following pages show screenshots from different components of the GUI and demonstrate the funconality of the GUI, following the workflow described above: • Main Menu • Creang Scenarios • Viewing and Eding Inputs • Running the Model • Reviewing Results

Main Menu: GUI on Opening The GUI is a macro enabled spreadsheet (an .xlsm file) and opens to a simple Main Menu Model selection drop down Scenario selection drop down Buttons for navigation within the spreadsheet, scenario management, and running the model Scenario inputs

Creang Scenarios

Creang a Scenario The scenario drop down list on the Main Menu shows the current list of scenarios for the selected model. At first, just the Base scenario exists

Creang a Scenario The model user names their scenario in the “Scenario Name” box. They can also specify which exisng scenario to use as a template for the new scenario using the “Copy From” drop down. In this case, the Base scenario will be copied and, unl the user edits the inputs to their new scenario, it will be idencal to the Base scenario. Clicking the “Create Scenario” button launches the Create Scenario Dialog.

Creang a Scenario The model user then clicks “okay” to create the new scenario. It is added to the scenario drop down as shown.

Creang a Scenario In addion, the new scenario folder is also added to the file system, in this case to \East Coast\LowAirFare.

Creang a scenario The file structure within the new scenario folder is idencal to that within the Base scenario folder (shown earlier in the user’s guide).

Eding Inputs

Eding Inputs The GUI provides access to two types of scenario inputs • Main scenario assumpons that can be set on the Main Menu tab • More detailed inputs that can viewed and edited on the Inputs tab These are described in turn on the following pages. It is these edits to inputs that allow a model user to disnguish their new scenario from the base and other scenarios that they have already evaluated.

Eding Main Scenario Assumpons Select the new scenario Edit the factors on the main menu The GUI contains 7 scenario input factors on the main menu. These allow the model user to make system level changes quickly in order to create the inputs for new scenarios quickly.

Editing Main Scenario Assumptions The scenario inputs factors are all system wide effects that factor up or down proporonally a parcular input. For example, setng Rail Fare to a value higher than 1, such as 1.2, increases all rail fares by that factor. So 1.2 mulplies all rail fares by 1.2, resulng in a 20% increase. Parameter Descripon Rail IVTT Factor on rail in vehicle travel me (changes system wide travel mes) Air IVTT Factor on air in vehicle travel me (changes system wide travel mes) Auto IVTT Factor on auto in vehicle travel me (changes air and rail access travel mes) Rail Fare Factor on rail fares (changes rail fares system wide) Air Fare Factor on air fares (changes air fares system wide) Rail Service Factor on number of trains per day (system wide) Air Service Factor on number of flights per day (system wide)

Eding Inputs The GUI also includes the capability to edit the more detailed input files to a scenario, specifically the various inputs describing the rail services. Select the new scenario Select “VIEW/EDIT INPUTS” button to go to the Inputs sheet

Eding Inputs The Inputs sheet is a standard interface that allows the model user to load inputs for the scenario into memory and then view and edit specific inputs using dialog boxes and the spreadsheet. Select the new scenario Select “VIEW/EDIT INPUTS” button to go the Inputs sheet The edit inputs interface loads as a blank spreadsheet; clicking on the “EDIT INPUTS” button reads in the data and starts the eding process

Eding Inputs Once the Edit Inputs buon has been clicked, a wizard takes the model user through the process of picking which parcular input they wish to view or edit Select the new scenario Select “VIEW/EDIT INPUTS” button to go the Inputs sheet Pick the specific input to edit or view

Eding Inputs The model user can choose parcular stations to appear at the upper le of the viewing/eding matrix, so they are easy to find. Select the new scenario Select “VIEW/EDIT INPUTS” button to go the Inputs sheet Then select staons of interest and they will appear at the upper le of the matrix

Eding inputs Model users can use the buttons and drop downs to switch to different inputs or exit if they have finishing viewing files. Select the new scenario Use the drop down menu to switch to a different input Click exit to leave the input editor

Eding inputs Once the data are loaded, the model user can edit them directly on the spreadsheet and then save the changes using the SAVE CHANGES button. The model user can then return to the main menu by clicking on MAIN MENU. The model user can edit data on the spreadsheet Once edits are made they can be saved to the input files for this scenario

Running a Scenario

Running a Scenario Once any file eding is done, the model user can run a scenario. A model run is started by clicking on the “RUN MODEL” buon. Select “RUN MODEL” button to run the current scenario

Running a scenario When the model user clicks on the “RUN MODEL” button, several steps are iniated: 1. Creaon of the “RunSpecificaon.txt” file, which describes the model and scenario to run. 2. R is launched via a command shell and the main R script containing the model applicaon code, ModelApplicaon.R, is executed. 3. The simulaon is carried out with logging of progress to a log file (“log.txt”).

Running a scenario During a scenario run the command shell window also shows and prints progress to the screen so that the user can see how the simulaon is progressing. Once command shell window closes, the run is complete and the mode user can processed to view the results.

Viewing Results

Reviewing Results At the end of a run the results are written into the Base and Future folders within the scenario. The model writes a complete set of outputs including the full simulaon results into an R binary file and exports key summary tabulaons into .csv files

Reviewing Results Click on the “VIEW OUTPUTS” button on the main menu to navigate to the outputs page. Select “VIEW OUTPUTS” button to go to the outputs sheet

Reviewing Results The outputs page allows the user to open results from any scenario that has been run, to view base and future results, and to view various tabulaons and charts. First, select a Scenario and Year using the drop down menus at the top of the page

Reviewing Results Once the user selects a table or chart to view the results are loaded and the data are displayed. Addional results including very detailed outputs are available via the file system. Pick a table from the tables drop down, or click on “VIEW CHARTS” and pick a chart from the charts drop down

Input File Specificaons

Input File Specificaons This secon of the user’s guides provides a complete descripon of the input files for the model. As noted above, the model comes with two pre developed models, for the East Coast and West Coast regions, which are contained in the “East Coast” and “West Coast” folders, respecvely. Within each of those two folders are a list of scenarios. Inially, there is only a “Base” scenario folder, but as scenarios are created, addional scenario folders will appear here. The screenshot below shows the contents of the “East Coast” folder – just the “Base” scenario folder when the model is installed.

Model Files: Inputs and Outputs Each scenario folder for example, within the East Coast/Base folder as shown here-- contains folders containing .csv files for several different categories of inputs, and the “Base” and “Future” folder which contain outputs (once the scenario has been run). The ScenarioParameters.csv file contains several top level policy variables and is edited using the GUI’s main menu.

Model Files: _Geography The geography folder contains one file, Counes.csv. This file describes the extent of the study area for this model and is simply a list of county and state names. The first 10 rows are shown in the table below. NAME STATE_NAME fairfield conneccut harord conneccut litchfield conneccut middlesex conneccut new haven conneccut new london conneccut tolland conneccut windham conneccut kent delaware new castle delaware Counes.csv Describes the extent of the study area Field Descripons: NAME – county name in text STATE_NAME – state name in text

Model Files: 1_Socioeconomic Data (1) The 1_Socioeconomic Data folder contains 5 files describing total air and rail travel and the demographics of travelers and more detailed spa
al informa
on about the study area. CountyOd_Roundtrips_Base.csv Describes exis
ng OD travel in the study area Field Descrip
ons: orgfips – FIPS code for the origin county desfips – FIPS code for the des
na
on county air_trips – annual air passenger trips in 2008 between the coun
es rail_trips – annual rail passenger trips in 2008 between the coun
es GCD – great circle distance between the coun
es hwydist – distance over the highway network between the coun
es orgfips desfips air_trips rail_trips GCD hwydist 9001 9001 0 13 0 23.8 9001 9003 0 1142 50.8 88.6 9001 9005 0 96 38.1 74.4 9001 9007 0 223 45.4 82.6 9001 9009 0 1209 24.8 65.8 9001 9011 0 810 67.6 94.4 9001 9013 0 276 68 102.6 9001 9015 0 124 82.8 131.5 9001 10001 8 47 187.3 276.6 9001 10003 67 2636 158.5 228.1

Model Files: 1_Socioeconomic Data (2) CountyOd_Roundtrips_Future.csv Describes future year OD travel in the study area Field Descrip
ons: orgfips – FIPS code for the origin county desfips – FIPS code for the des
na
on county air_trips – annual air passenger trips in 2040 between the coun
es rail_trips – annual rail passenger trips in 2040 between the coun
es GCD – great circle distance between the coun
es hwydist – distance over the highway network between the coun
es orgfips desfips air_trips rail_trips GCD hwydist 9001 9001 0 18 0 23.8 9001 9003 0 1516 50.8 88.6 9001 9005 0 132 38.1 74.4 9001 9007 0 317 45.4 82.6 9001 9009 0 1608 24.8 65.8 9001 9011 0 1097 67.6 94.4 9001 9013 0 404 68 102.6 9001 9015 0 179 82.8 131.5 9001 10001 8 76 187.3 276.6 9001 10003 64 3700 158.5 228.1

Model Files: 1_Socioeconomic Data (3) PartyDemographics.csv Describes the joint income, vehicle availability and trip purpose distribu
ons for air and rail travel par
es Field Descrip
ons: Purpose – trip purpose groups (Business and NonBusiness) Income – household income groups (<$25,000, $25,000 to $75,000, $75,000 to $125,000, and $125,000 or more) VehicleAvailability – whether the travel party has a vehicle available or not (No, Yes) AirProp – the propor
on of air travel par
es in each joint category of Purpose, Income, and Vehicle Availability. Sums to 1.0 RailProp – the propor
on of rail travel par
es in each joint category of Purpose, Income, and Vehicle Availability. Sums to 1.0 Purpose Income VehicleAvailability AirProp RailProp Business $125,000 or more No 0 0 Business $125,000 or more Yes 0.22 0.12 Business $75,000 to $125,000 No 0 0 Business $75,000 to $125,000 Yes 0.11 0.18 Business $25,000 to $75,000 No 0 0 Business $25,000 to $75,000 Yes 0.09 0.09 Business <$25,000 No 0 0 Business <$25,000 Yes 0.01 0 NonBusiness $125,000 or more No 0 0 NonBusiness $125,000 or more Yes 0.17 0.07 NonBusiness $75,000 to $125,000 No 0 0.01 NonBusiness $75,000 to $125,000 Yes 0.13 0.21 NonBusiness $25,000 to $75,000 No 0.01 0.01 NonBusiness $25,000 to $75,000 Yes 0.21 0.25 NonBusiness <$25,000 No 0.01 0 NonBusiness <$25,000 Yes 0.03 0.06

Model Files: 1_Socioeconomic Data (4) PartySize.csv Describes the party size distribuon for air and rail travel pares by trip purpose Field Descripons: Purpose – trip purpose groups (Business and NonBusiness) PartySize – party size of the travel party (1, 2, 3, 4, 5 or more) AirProp – the proporon of air travel pares in each party size category. Sums to 1.0 for each purpose group, 2.0 overall. RailProp – the proporon of rail travel pares in each party size category. Sums to 1.0 for each purpose group, 2.0 overall. Purpose PartySize AirProp RailProp Business 1 0.8 0.75 Business 2 0.14 0.17 Business 3 0.03 0.05 Business 4 0.02 0.02 Business 5 or more 0.02 0.02 NonBusiness 1 0.49 0.29 NonBusiness 2 0.37 0.34 NonBusiness 3 0.06 0.13 NonBusiness 4 0.04 0.15 NonBusiness 5 or more 0.04 0.1

Model Files: 1_Socioeconomic Data (5) TractData.csv Describes the characteriscs of each Census tract in the study area Field Descripons: TractNum – Census Bureau tract idenficaon number TractID – sequenal tract idenficaon number used in the model, 1:n tracts CtyNum – county FIPS code CtyName – text county name StNum – state FIPS code StName – text state name Pop – tract populaon HHMedInc – tract household median income ($) AveHHVeh – tract average household vehicle availability Emp_Tot – tract total employment Emp_Hosp – tract employment in the hospitality industry TractNum TractID CtyNum CtyName StNum StName Pop HHMedInc AvgHHVeh Emp_Tot Emp_Hosp 9001010101 1 9001 fairfield county 9 conneccut 4476 154421 2.342049 2913 147 9001010102 2 9001 fairfield county 9 conneccut 4330 241944 2.293716 913 13 9001010201 3 9001 fairfield county 9 conneccut 3421 250001 2.537137 839 10 9001010202 4 9001 fairfield county 9 conneccut 5359 175625 2.202524 414 0 9001010300 5 9001 fairfield county 9 conneccut 4010 152321 2.084577 4952 191 9001010400 6 9001 fairfield county 9 conneccut 5290 83036 1.78829 2198 170 9001010500 7 9001 fairfield county 9 conneccut 5494 88750 1.548553 2848 129 9001010600 8 9001 fairfield county 9 conneccut 1845 90909 1.331522 10852 699 9001010700 9 9001 fairfield county 9 conneccut 3579 65682 1.34878 336 7 9001010800 10 9001 fairfield county 9 conneccut 3388 135204 1.885401 1230 150

Model Files: 2_RailStaon Access Data The 2_RailStaon Access Data folder contains one file, RailAccess.csv. This file describes the accessibility of each station to each tract in the study area. RailAccess.csv Describes the accessibility of each rail staon to each tract in the study area Field Descripons: RailAccInd – unique idenfier for the combinaon of tract and staon, 1:n TractInd – sequenal tract idenficaon number used in the model, 1:n tracts StaonInd sequenal rail staon idenficaon number used in the model, 1:n staons AutoTT – auto travel me from tract to staon HwyDist – highway distance from tract to staon Transit – transit accessibility index, tract staon combinaon RailAccInd TractInd StaonInd AutoTT HwyDist Transit 1 1 1 256.38 270.59 0.011 2 1 2 251.63 262.24 0.008 3 1 3 247.9 258.44 0.008 4 1 4 233.1 243.79 0.01 5 1 5 222.25 228.7 0.012 6 1 6 223.4 224.22 0.008 7 1 7 209.77 212.32 0.008 8 1 8 188.21 194.96 0.024 9 1 9 173.29 178.7 0.018 10 1 10 169.36 178.2 0.008

Model Files: 3_Airport Access Data The 3_Airport Access Data folder contains one file, AirportAccess.csv. This file describes the accessibility of each airport to each tract in the study area. AirAccess.csv Describes the accessibility of each airport staon to each tract in the study area Field Descripons: AccInd – unique idenfier for the combinaon of tract and airport, 1:n TractInd – sequenal tract idenficaon number used in the model, 1:n tracts AirportInd sequenal airport idenficaon number used in the model, 1:n airports AutoTT – auto travel me from tract to airport HwyDist – highway distance from tract to airport Transit – transit accessibility index, tract airport combinaon AccInd TractInd AirportInd AutoTT HwyDist Transit 1 1 1 51.54 50.31 0.014 2 1 2 96.52 94.39 0.008 3 1 3 245.38 253.24 0.008 4 1 4 260.57 270.68 0.008 5 1 5 174.11 172.43 0.011 6 1 6 171.36 180.94 0.008 7 1 7 207.76 209.67 0.008 8 1 8 330.13 246.89 0.008 9 1 9 263.96 255.02 0.008 10 1 10 242.49 219.75 0.011

Model Files: 4_Rail Travel Data (1) The 4_Rail Travel Data folder contains three files and two subfolders that contain files describing the rail station loca
ons and rail service between them. The Skims folder contains a single file, skims.csv. The SkimWeights folder contains a set of files used by the model to support adjustments to the station to station travel
mes and should not be edited by the model user.

Model Files: 4_Rail Travel Data (2) RailStaon.csv Describes the locaon of rail staons in the study area Field Descripons: StaonInd – sequenal staon idenficaon number used in the model, 1:n staons StaonCod – 3 le‚er code for the staon StaonName – text name for the staon TractID – sequenal tract idenficaon number used in the model, 1:n tracts; this is the tract containing the staon NAME – Name of the county the staon is located in STATE_NAME – Name of the state the staon is located in Latude – latude of the staon Longitude – longitude of the staon KeyStaon * idenfying major staons (e.g., South Boston Staon) StaonInd StaonCod StaonName TractID NAME STATE_NAME Latude Longitude KeyStaon 1POR Portland, ME (POR) 1257Cumberland Maine 43.63549 70.2949 2ORB Orchard, ME (ORB) 1303York Maine 43.51731 70.3776 3 SAO Saco, ME (SAO) 1299York Maine 43.50092 70.4428 4WEM Wells, ME (WEM) 1330York Maine 43.32203 70.5809 5DOV Dover, NH (DOV) 4353Strafford New Hampshire 43.19786 70.8737 6DHM Durham, NH (DHM) 4347Strafford New Hampshire 43.13397 70.9265 7 EXR Exeter, NH (EXR) 4337Rockingham New Hampshire 42.98148 70.9478 8HHL Haverhill, MA (HHL) 3073Essex Massachuse‚s 42.7762 71.0773 9WOB Woburn, MA (WOB) 3355Middlesex Massachuse‚s 42.47926 71.1523 10BON N Bost, MA (BON) 3853Suffolk Massachuse‚s 42.36558 71.0613 11BOS S Bost, MA (BOS) 3887Suffolk Massachuse‚s 42.35152 71.0553*

Model Files: 4_Rail Travel Data (3) Skims.csv Describes the rail level of service between staon pairs in the study area (each of these fields can be edited in the GUI as described earlier in the user’s guide) Field Descripons: From – StaonInd (staon idenficaon number used in the model, 1:n staons) for the origin staon To – StaonInd (staon idenficaon number used in the model, 1:n staons) for the desnaon staon Time – average rail travel me (minutes) Speed – average rail speed (mph) Cost – average cket cost ($) Distance – average rail distance (miles) Regional IVTT – regional rail service in vehicle travel me (minutes) HSR IVTT – high speed rail in vehicle travel me (minutes) Regional Speed – regional rail service average speed (mph) HSR Speed – high speed rail average speed (mph) Regional Layover – time transferring on regional rail services (minutes) HSR Layover me transferring on high speed rail services (minutes) Regional Cost – average cket cost on regional rail service ($) HSR Cost – average cket cost on high speed rail service ($) Regional Distance – distance on regional rail service (miles) HSR Distance – distance on high speed rail service (miles) Transfers – average number of transfers for service between origin and desnaon staon nPaths – number of alternave ineraries per week nRegPaths – number of alternave regional rail service ineraries per week nHSRPaths – number of alternave high speed rail service ineraries per week

Model Files: 4_Rail Travel Data (4) RailSchedule.csv Matrix describing each train service and its schedule, with one column describing the schedule for each service. This file is processed into Skims.csv by the model in conjunc­on with RailSta­onDistances.csv Field and row Descrip­ons: Mode (row) – code for service type, 1 Regional Rail, 2 Exis­ng Higher Speed Rail, 3 High Speed Rail, 4 Bus Distance Group (row) – code to iden­fy service groups (e.g., Northeast Regional service) Bound (row) – direc­on of travel (N,S) TrainName (row) name of the train service (Amtrak) TrainNumber (row) Amtrak train number Opera­on (row) – string describing the days the train operates (e.g., 1_2_3_4_5_6_7 operates every day, 6_7 operates on weekends only) Sta­onNum (column) Sta­onInd (sta­on iden­fica­on number used in the model, 1:n sta­ons) for the sta­on Sta­onCod 3 le•er code for the sta­on Sta­onName – text name for the sta­on Cell values: Each cell in the table encodes informa­on about the rail service, including sequence for that train, whether passengers can board, alight, or both, and the ­me of arrival and departure 01 R 0930P—0 01 indicates that this is the first sta­on for this train R indicates boarding only (B is both boarding and aligh­ng, D is aligh­ng only) 0930P is the arrival ­me at the sta­on, 9:30 PM 0 is the ­me that that sta­on stops at the sta­on, i.e., the scheduled departure ­me at the sta­on is also 9:30 PM

Model Files: 4_Rail Travel Data (5) RailStaonDistances.csv Matrix describing the staon to staon distances for each train service group, with one column describing the distances for each service group. This file is processed into Skims.csv by the model in conjuncon with RailSchedule.csv Field and row Descripons: Service (row) – name of the train service group (services with similar stopping paƒern) Group (row) – code to idenfy service groups StaonNum (column) StaonInd (staon idenficaon number used in the model, 1:n staons) for the staon StaonCode 3 leƒer code for the staon StaonName – text name for the staon Cell values: Each cell value represents the train distance travelled for that service. The value is 0 at the first staon and increases with each subsequent staon in the sequence of stops Service Northeast Regional SB StaonInd StaonCode StaonName Group 1 1POR Portland, ME 2ORB Old Orchard Beach, ME 3SAO Saco, ME 4WEM Wells, ME 5DOV Dover, NH 6DHM Durham UNH, NH 7EXR Exeter, NH 8HHL Haverhill, MA 9WOB Woburn, MA 10BON Boston, MA North Staon 11BOS Boston, MA South Staon 0 12BBY Boston, MA Back Bay Staon 1 13RTE Route 128, MA 11 14PVD Providence, RI 43 15KIN Kingston, RI 70 16WLY Westerly, RI 87 17MYS Mysc, CT 96 18NLC New London, CT (Casino) 105 19OSB Old Saybrook, CT 123

Model Files: 5_Air Travel Data (1) The 5_Air Travel Data folder contains two files that describe airport locaon and air service.

Model Files: 5_Air Travel Data (2) Airport.csv Describes the locaon of airports in the study area Field Descripons: AirportInd – sequenal airport idenficaon number used in the model, 1:n airports AirportCod – 3 le€er code for the airport AirportName – text name for the airport TractID – sequenal tract idenficaon number used in the model, 1:n tracts; this is the tract containing the airport NAME – Name of the county the airport is located in STATE_NAME – Name of the state the airport is located in Latude – latude of the airport Longitude – longitude of the airport AirportInd AirportCod AirportName TractID NAME STATE_NAME Latude Longitude 1HVN TWEED NEW HAVEN 634New Haven CONNECTICUT 41.26375 72.8868 2BDL BRADLEY INTL 435Haršord CONNECTICUT 41.93889 72.6832 3DCA RONALD REAGAN WASHINGTON NATIONAL 11888Arlington DIST. OF COLUMBIA 38.85208 77.0377 4 IAD WASHINGTON DULLES INTL 12486 Loudoun DIST. OF COLUMBIA 38.94744 77.4599 5BED LAURENCE G HANSCOM FLD 3489Middlesex MASSACHUSETTS 42.46995 71.289 6BOS GENERAL EDWARD LAWRENCE LOGAN INTL 4014Suffolk MASSACHUSETTS 42.36297 71.0064 7HYA BARNSTABLE MUNI BOARDMAN/POLANDO FIELD 2770Barnstable MASSACHUSETTS 41.66933 70.2804 8ACK NANTUCKET MEMORIAL 3583Nantucket MASSACHUSETTS 41.25311 70.0603 9PVC PROVINCETOWNMUNI 2715Barnstable MASSACHUSETTS 42.07228 70.2207 10MVY MARTHAS VINEYARD 2940Dukes MASSACHUSETTS 41.39342 70.6139

Model Files: 5_Air Travel Data (3) Airport.csv Describes the locaon of airports in the study area Field Descripons: ORIGIN – origin airport code (3 le er) DEST – desnaon airport code (3 le er) Numpersons – number of persons traveling between airports (from DB1B) N_direct – number of persons traveling direct between the two airports (from DB1B) N_transfer – number of persons traveling between the airports with one or more transfers in between (from DB1B) N_oneway – number of one way ineraries (from DB1B) N_round – number of round trip ineraries (from DB1B) Numdays – number of day reported for this airport pair in the on me performance data (OTP) Numflights – number of flights (OTP) Cancelled number of cancelled flights (OTP) Dep_delay – average departure delay in minutes (OTP) Arr_delay – average arrival delay in minutes (OTP) Sche_me – scheduled travel me in minutes (OTP) Actual_me actual travel me in minutes (OTP) Distance distance between airports (OTP) Tran_days – days reported for transfer flight routes in the OTP Tran_flight – number of transfer flights reported (OTP) Tran_cancelled – number of transfer flights cancelled (OTP) Tran_flight_me – tranfer flight travel me in minutes (OTP) Tran_layover – layover me in minutes (OTP) Mean_cost – average cket cost (from DB1B, in $) p0:p100 – 0 percenle to 100 percenle cket costs in 10 percenle increments (from DB1B, in $)

Code Guide

Code Guide This secon of the user’s guide provides a guide to the code of the model and the files containing model parameters. The informaon in this secon of the user’s guide is intended for advanced users of the model who wish to modify the model – from making adjustments to individual parameters, to making fundamental changes to the structure of individual model components or the overall model framework.

R Resources The R applicaon comes packaged with the ACRP 3 23 Air/Rail Diversion Model. To work with the R scripts and R binary files outside of the model’s GUI, using an integrated development environment (IDE), such as R Studio, is recommended. R Studio is an open source applicaon and can be downloaded at h†p://www.rstudio.com/. For new R users, model users are recommended to familiarize themselves with the R language prior to a†empng to edit the model’s R scripts. Resources such as h†p://www.r project.org/ and h†p://www.r bloggers.com/ are good places to start.

Model Files: R Model, Applicaon The _RModel folder contains the R folder, which is a version of the R applicaon that is used to run the model. The files in that folder are as follows: • ACRPData_.Rdata – R binary folder containing several input tabulaons. • ACRPModel_.Rdata – R binary folder containing model coefficients. • Log.txt – log file from a model run (added when the model is run in installaon). • ModelApplicaon.R – R script containing the model code. • RunModel.R – R script to run the model (added when the model is run in installaon). • RunSpecificaon.txt – text file written by the Excel GUI providing variables to R (added when the model is run in installaon).

Model Files: ModelApplicaon.R (1) ModelApplicaon.R is the R script containing the model code and encompasses all of the R code to run the model. Eding this script will allow the structure of individual model components or the overall model structure to be changed. The overall structure of the file is as follows: • Header (Lines 1 14): file meta data (e.g., author, version), and code to begin logging and ming. • Funcons (Lines 15 58): defined funcons used later in the script to load R packages and to simulate logit models give a data set and model specificaon/coefficients. • Libraries (Line 59 67): load R packages required by the model (note: R packages are separate libraries of addional funcons). • Directories, input files, output files, models (Lines 68 226): loads in the input data, model coefficients, and creates directories for the outputs. • Make a list to store the directory references (Lines 73 111): creates a list structure to hold the directory references for inputs and outputs. • Read in input data (Lines 112 – 167): reads into memory all of the input data for the model being executed. • Load the tabulaons of model coefficients (Lines 168 200): loads the model coefficients and applies calibraon adjustments to them.

Model Files: ModelApplicaon.R (2) • Directories, input files, output files, models (Lines 68 226): loads in the input data, model coefficients, and creates directories for the outputs • Load validaon/general correspondence data (Lines 201 226): loads in tables of correspondences, and also applies the parameters from the run specificaon to the inputs data. • Model Simulaon (Lines 228 881): The simulaon applies the 6 model steps. • Line 243 start the loop on years: the model is sequenally applied first for the base year and then for the future year. This loop runs unl line 1031 and includes the next main secon of code from line 882 to 1031 where summary model outputs are produced. • 1. Travel party sampling from enumerated list of travel pares (Lines 247 313) produces a travel party sampling from enumerated list of travel pares. • 2. Allocaon of travel pares to Census Tracts (Lines 314 338) assigns each travel party to a census origin and desnaon tract. • 3. Simulaon of addional travel party characteriscs (Lines 339 354) adds income category and vehicle availability for each travel party. • 4. Airport and station choice and access and egress mode choice model (Lines 355 661) applies in turn the airport and station choice models: • Calculate airport to airport ulity for all of the airport pairs with air service (Lines 360 402).

Model Files: ModelApplicaon.R (3) • Model Simulaon (Lines 228 881): The simulaon applies the 6 model steps. • 4. Airport and station choice and access and egress mode choice model (Lines 355 661) applies in turn the airport and station choice models: • Calculate access and egress ulies for each airport tract combinaon (Lines 403 413). • Builds the choiceset of otract oairport dairport dtract combinaons for each otract dtract combinaon (Lines 414 442). • Apply the choice model for each party in the SynPop.. (Lines 443 490) • Station Choice (Lines 491 574). • HSR Station Choice (Lines 575 660). • 5. Main mode choice model (Lines 661 783) applies the main mode choice model between air and rail. • 6. Airline Response Model (Lines 784 871) applies the airline response model including reapplying the main mode choice model. • Save large model output files (Lines 872 876) . Save R binary files containing the detailed tables of results for each travel party. • Produce Summary Model Outputs (Lines 882 1031) writes summary results. • County Summaries (Lines 896 960) are produced in memory. • Airport and Station summaries (Lines 961 990) are produced in memory. • Airline Response Summaries (Lines 991 1001) are produced in memory. • Results, including summaries for GUI are written to file (Line 1001 1031).

Model Files: ACRPData_.Rdata ACRPData_.Rdata is an R binary folder containing several input tabulaons that include geographical correspondence and validaon data. It is loaded into memory in the Load validaon/general correspondence data secon of the simulaon at line 203: • Apeast is a table of observed passenger trips in 2008 between airports in the East coast study area. • Apwest is a table of observed passenger trips in 2008 between airports in the West coast study area • Countymsa is a correspondence table between counes and metropolitan statiscal areas. • Steast is a table of observed passenger trips in 2008 between rail stations in the east coast study area. • Stwest is a table of observed passenger trips in 2008 between rail staons in the west coast study area.

Model Files: R Model, Applicaon ACRPModel_.Rdata is an R binary folder containing model coefficients. It is loaded into memory in the Load the tabulaons of model coefficients secon of the simulaon at line 170. The file contains an R list object that contains 5 elements: • MainModeChoiceBusiness: a table of coefficients for the business segment in the main mode choice model • MainModeChoiceNonBusiness: a table of coefficients for the non business segment in the main mode choice model • AirportAccessChoice: a table of coefficients for the airport choice model • CensusTractAllocaon: a list of coefficients of the census tract allocaon model • AirlineResponse: a table of coefficients for the airline response model