Appendix B-1 Technical Memoranda - New York State ......NEW YORK STATE DEPARTMENT OF TRANSPORTATION KOSCIUSZKO BRIDGE PROJECT ... Flushing/Bedford Rezoning FEIS (March 16, 2001) TRANSCOM

Appendix B-1 Technical Memoranda

Technical Memorandum T1: Traffic Data Collection Plan

Technical Memorandum T3: Travel Demand/Traffic Simulation Modeling Methodology

May 11, 2005 Page 1

NEW YORK STATE DEPARTMENT OF TRANSPORTATION KOSCIUSZKO BRIDGE PROJECT

TECHNICAL MEMORANDUM T-1

TRAFFIC DATA COLLECTION PLAN 1.0 INTRODUCTION This technical memorandum presents the process by which the New York State Department of Transportation (NYSDOT) will collect data for the traffic analysis to be conducted as part of the Kosciuszko Bridge Project. This data will be used to establish the base year traffic conditions within the limits of the project and will be utilized to model future traffic conditions and evaluate project alternatives. The traffic data collection process consists of three primary steps. First, contact is made with all agencies that could potentially have relevant traffic data within the project’s traffic study area, shown in Figure 1. Second, the available data is reviewed for validity and applicability to the project and deficiencies in the overall data identified. Finally, additional data is collected to complete the data needs for the project. The following sections detail this process as applied for the Kosciuszko Bridge Project. The data collection effort for the Kosciuszko Bridge Project is based on information derived from NYSDOT’s Kosciuszko Bridge Traffic Operations Study completed in November 1995. The data and findings from that study were used to guide the selection of data collection locations for this project. 2.0 COLLECTION OF EXISTING DATA In addition to NYSDOT, a number of agencies throughout the region collect traffic data as part of various studies and projects. The following agencies and offices were contacted for data that may be relevant to the project:

• Automobile Club of New York • Brooklyn Borough President’s Office • Brooklyn Community Board #1 • Long Island City Business Development

Corporation • Metropolitan Transportation Authority

(MTA) o MTA Bridges and Tunnels –

Revenue Management o MTA New York City Transit

• New York City Department of City Planning (NYCDCP)

o Brooklyn Borough Office o Environmental Assessment &

Review o Queens Borough Office o Transportation Division

• New York City Department of Sanitation • New York City Department of

Transportation (NYCDOT) o Brooklyn Division

o Queens Division o Signals o Traffic Data o Traffic Operations o Transit Operations

• New York City Police Department (NYPD) • New York Metropolitan Transportation

Council (NYMTC) • New York State Motor Truck Association • New York State Movers and

Warehousemen • North Brooklyn Development Corporation • Obusty Local Development Corporation • Port Authority of New York and New

Jersey – Office of Policy and Planning • Queens Borough President’s Office • Queens Community Board #2 • Queens Community Board #5 • SIMCO Engineering, P.C. • St. Nicholas Preservation Development

Corporation

Kosciuszko Bridge Project – Technical Memorandum T-1

May 11, 2005 Page 2

• Transportation Alternatives • Transportation Operations Coordinating

Committee (TRANSCOM)

• West Maspeth Local Development Corporation

The majority of these agencies had no data that was relevant to the project. Table 1 lists the agencies with relevant data and the data that was provided. TABLE 1: LIST OF AVAILABLE DATA FROM AGENCIES Agency Available Data NYSDOT 1 – ATR counts in study area 2 – Turning Movement Counts in study area 3 – Classification counts 4 – Transportation Improvement Program (TIP) 5 – Highway sufficiency rating 6 – Cross Harbor Tunnel DEIS NYCDOT 1 – Bridge traffic volumes (1999 & 2000) 2 – Seasonal traffic flow factors (1999 & 2000) 3 – New York City cycling map 4 – Bus timetable for Triboro Coach and Queens Surface Corp. 5 – Signal plans for critical intersections in study area 6 – Yearly bridge opening reports (1999, 2000, & 2001) 7 – Truck route map within study area 8 – New York City Waterways Project (December 1999) NYCDOS Comprehensive Solid Waste Management Plan Modification FEIS (2000) NYCDCP 1 – Citywide Congestion Bottleneck Study, Phase 2 (August 2001) 2 – Flushing/Bedford Rezoning FEIS (March 16, 2001) TRANSCOM Ongoing and long-term projects in New York MTA-NYCT Grand Avenue Depot and Central Maintenance Facility II, FEA (March 2001) Brooklyn Navy Yard Development Corporation (BNYDC)

Brooklyn Navy Yard Traffic Impact Study (July 1998)

3.0 REVIEW OF EXISTING DATA The traffic data received from the above agencies was reviewed for validity and relevance to the project. Most of the data was collected prior to Fall 1999 and therefore does not satisfy the requirements of the project. This data will be used for comparative purposes only. Tables 2, 3, and 4 and Figures 2, 3, and 4 describe and show, respectively, the automatic traffic recorder (ATR), turning movement counts (TMC) and classification counts currently available. 4.0 SUPPLEMENTAL DATA COLLECTION After reviewing the available data, a supplemental data collection plan was developed to provide adequate data for the project’s traffic analysis. This plan, as described on the following pages, proposes to collect the following additional data:

• Traffic flow volumes • Turning movement counts • Vehicle classification counts • Occupancy counts • Travel time and delay runs

• Pedestrian crosswalk counts • Physical inventory survey • On-street parking survey • Accident data


May 11, 2005 Page 3

4.1 Traffic Flow Volumes Twenty-four hour ATR counts are proposed at 74 locations within the project study limits. Each ATR will be placed mid-block and collect data continuously for a period of one week. This will provide data on traffic flow variations, average daily traffic (ADT) volumes, and ultimately will be used to develop the balanced traffic volume network for the project. Each ATR count will be calibrated to convert axle counts to total vehicles. The proposed ATR count locations are described in Table 5 and shown in Figure 5. 4.2 Turning Movement Counts Intersection TMCs are proposed at 57 locations. This data will be collected concurrently with the ATR counts during the morning (6:00 – 10:00 AM), midday (11:00 AM – 2:00 PM) and afternoon (3:00 – 7:00 PM) peak periods on one mid-weekday (i.e. Tuesday, Wednesday, or Thursday) and on a Saturday during the midday peak period (10:00 AM – 3:00 PM). The data will be collected in 15-minute intervals and vehicles will be grouped into one of three categories by general type: cars, buses, and trucks. The proposed TMC locations are described in Table 6 and shown in Figure 6. 4.3 Vehicle Classification Counts Detailed classification counts will be performed at the nine locations listed in Table 7 and shown in Figure 7. Counts will be performed during the same weekday morning, midday, and afternoon and Saturday midday peak periods described in Section 4.2. The counts will collect data for five categories of vehicles:

• Automobiles • Buses • Light-duty gasoline trucks (2 axles) • Heavy-duty gasoline trucks (3 or more axles) • Heavy-duty diesel trucks (3 or more axles)

4.4 Occupancy Counts Occupancy counts will be performed concurrently with the vehicle classifications counts described in Section 4.3 at the locations listed in Table 7 and shown on Figure 7. Occupancy data will be recorded for passenger vehicles in the following groups: (1) one, (2) two, (3) three and (4) four or more occupants. 4.5 Travel Time and Delay Runs Travel time and delay runs will be conducted on 15 major routes in the study area using the “floating car” method (the test car attempts to “float” in the traffic stream to obtain a representative travel time). Three runs will be done in each direction during the same weekday morning, midday, and afternoon and Saturday midday peak periods as described in Section 4.2. Elapsed time and any en route delays (e.g. accident, signal, vehicle breakdown, etc.) will be recorded for each of the segments along each route. The proposed routes for travel time and delay runs are described in Table 8 and shown in Figure 8. 4.6 Pedestrian Crosswalk Counts Pedestrian crosswalk counts will be performed at the 10 intersections listed in Table 9 and shown in Figure 9. These counts will be performed during the same weekday morning, midday, and afternoon and Saturday midday peak periods described in Section 4.2.


May 11, 2005 Page 4

4.7 Physical Inventory Survey A physical inventory survey will be conducted for each of the 15 major travel routes listed in Table 8 for travel time and delay runs and the 57 intersections listed in Table 5 for TMCs. Data on existing roadway geometry and traffic control regulations will be collected, including:

• Roadway, sidewalk and crosswalk widths • Number of travel lanes • Curb parking regulations • Lane utilization (proposed turning movement count intersections only) • Signal timing (proposed turning movement count intersections only) • Traffic control devices (type and location) • Bus stop locations, bus routes, and frequency of buses • Posted speed limits • Loading areas • Truck Routes • Off-street parking facilities • Direction of travel (one-way streets) • Locations of entrance and exit ramps

4.8 On-Street Parking Survey On-street parking will be surveyed during the same weekday morning, midday, and afternoon and Saturday midday peak periods described in Section 4.2. The survey will collect information regarding the number of vehicles parked along each roadway segment during each period. The roadways to be surveyed will include likely diversion routes and will be determined based on the construction staging plans for each alternative. 4.9 Accident Data Available traffic accident statistics will be obtained from NYSDOT’s Traffic and Safety Division State Accident Surveillance System (SASS) and Centralized Local Accident Surveillance System (CLASS) databases for the most recent 2-year period. Additionally, Department of Motor Vehicles Report of Motor Vehicle Accident Forms (MV104) will be obtained for the same period. Reported accidents on the BQE mainline and ramps within the project limits will be compiled and accident rates will be calculated for mainline accident occurrences per million vehicle kilometer (MVK) and for ramp accident occurrences per million entering vehicles (MEV). The computed accident rates for these major routes will be compared with the statewide average rates. On-site field reconnaissance trips will also be performed to obtain additional insight into possible accident causal factors and influences. Appropriate mitigation measures will be developed for high accident locations based on identification of dominant accident patterns and possible accident causal factors associated with non-standard geometric features and/or traffic operation characteristics. 5.0 NEXT STEPS Upon approval from NYSDOT, the data collection effort described above will be carried out. The collected data will be summarized as part of Technical Memorandum T-2: Baseline Traffic and Transportation Data Summary and will be used in the project’s traffic analysis.

TABLE 2: AVAILABLE ATR COUNTS

ID Roadway From To Start Date End DateData Source1

1 I-278 BQE Exit 30 Flushing Ave Exit 32 Williamsburg Bridge 3/16/1999 3/19/1999 NYSDOT 12 Meeker Ave Metropolitan Ave McGuinnes Blvd 8/12/2000 8/18/2000 NYSDOT 13 I-278 BQE Exit 32 Williamsburg Bridge Exit 33 McGuinnes Blvd 7/31/2001 8/3/2001 NYSDOT 1

5/25/1999 5/28/19994 I-278 BQE Exit 33 McGuinnes Blvd Koscuiszko Bridge 7/31/2001 8/3/2001 NYSDOT 1

5/25/1999 5/28/19995 Meeker Ave McGuinnes Blvd Porter Ave 7/30/2001 8/3/2001 NYSDOT 16 I-278 BQE Koscuiszko Bridge Exit 35 Rt I-495 5/14/2002 5/17/2002 NYSDOT 1

11/7/2000 11/10/200010/14/2001 10/19/20018/14/2001 8/17/2001

7 I-278 BQE Exit 35 Rt I-495 Exit 37 Roosevelt Ave 8/14/2001 8/17/2001 NYSDOT 18 Maurice Ave I-495 69th St 9/12/1999 9/17/1999 NYSDOT 19 Grand Ave Flushing Ave I-495 9/10/2001 9/14/2001 NYSDOT 1

6/8/1999 6/11/199910 I-495 Exit 17 WB I-278 Exit 19 Woodhaven Blvd 7/12/1999 7/16/1999 NYSDOT 111 I-495 Exit 14 21st St Exit 17 WB I-278 11/30/1999 12/7/1999 NYSDOT 1

3/29/1999 4/6/199912 I-495 Queens County Line Exit 14 21st St 10/22/2001 10/25/2001 NYSDOT 113 Jackson Ave Borden Ave 11th St 11/1/1999 11/5/1999 NYSDOT 114 Jackson Ave 21st St Queens Blvd 6/1/1999 6/6/1999 NYSDOT 115 Thompson Ave Jackson Ave Van Dam St 9/28/1999 10/1/1999 NYSDOT 116 Van Dam St I-495 Queens Blvd 3/2/1999 3/5/1999 NYSDOT 117 Van Dam St Greenpoint Ave I-495 8/13/2001 8/17/2001 NYSDOT 1

10/5/1999 10/10/199918 Greenpoint Ave Van Dam St 39th St 9/9/2001 9/14/2001 NYSDOT 119 Greenpoint Ave Kings County Line Van Dam St 11/27/2000 12/1/2000 NYSDOT 1

6/1/1999 6/6/199920 Jackson Ave 11th St 21st St 9/18/2000 9/22/2000 NYSDOT 1

11/1/1999 11/5/199921 McGuinnes Blvd Greenpoint Ave Queens County Line 8/12/2000 8/18/2000 NYSDOT 122 McGuinnes Blvd I-278 BQE Greenpoint Ave 8/12/2000 8/18/2000 NYSDOT 123 Bushwick Ave Montrose Ave Meresole St 7/29/2001 8/3/2001 NYSDOT 124 Metropolitan Ave I-278 BQE Bushwick Ave 10/1/2000 10/6/2000 NYSDOT 125 Metropolitan Ave Bushwick Ave Grand St 7/30/2001 8/3/2001 NYSDOT 126 Metropolitan Ave Vandervoort Ave Gardner Ave 4/27/1998 5/1/1998 NYCT27 Flushing Ave Kings County Line Metropolitan Ave 8/13/2001 8/17/2001 NYSDOT 128 Metropolitan Ave Kings County Line Eliot Ave 9/10/2001 9/17/2001 NYSDOT 129 Flushing Ave Metropolitan Ave Grand Ave 10/16/2000 10/20/2000 NYSDOT 130 Grand Ave Kings County Line Flushing Ave 11/27/2000 12/1/2000 NYSDOT 1

6/8/1999 6/11/199931 Grand Ave 47th St 49th St 11/2/1999 11/11/1999 NYCT32 Grand St Bridge Brooklyn County Line Queens County Line 10/26/1998 10/30/1998 NYCT33 Grand St Exit I-278 BQE Grand St 8/19/2000 8/25/2000 NYSDOT 134 Grand St Grand St Exit Metropolitan Ave 7/29/2001 8/3/2001 NYSDOT 135 69th St Grand Ave Maurice Ave 9/10/2001 9/14/2001 NYSDOT 1

6/8/1999 6/11/199936 69th St Maurice Ave Queens Blvd 6/8/1999 6/11/1999 NYSDOT 137 Grand Ave I-495 Queens Blvd 6/13/1999 6/18/1999 NYSDOT 138 56th Rd 48th St 44th St 11/13/2001 11/15/2001 NYSDOT 539 Flushing Ave Kent Ave Franklin Ave 6/15/1999 6/17/1999 NYCDCP 240 I-278 BQE Exit 29 Tillary Ave Exit 30 Flushing Ave 3/23/1998 3/30/1998 BNYDC41 Kent Ave Keap St Ross St 3/23/1998 3/30/1998 BNYDC

1 See Table 1 for description of project data sources.Note: See Figure 2 for Available ATR Counts

TABLE 3: AVAILABLE TURNING MOVEMENT COUNTS

ID Intersection Date of Count Data Source1

1 Vandervoort Ave @ Metropolitan Ave 11/28/2000 NYCT2 Vandervoort Ave @ Grand St 11/28/2000 NYCT3 Metropolitan Ave @ Gardner Ave 11/28/2000 NYCT4 Metropolitan Ave @ Woodward Ave 11/28/2000 NYCT5 Metropolitan Ave @ Flushing Ave 11/28/2000 NYCT6 Flushing Ave @ 54th St 11/28/2000 NYCT7 Flushing Ave @ 55th St 11/28/2000 NYCT8 Grand Ave @ 47th St 11/28/2000 NYCT9 Grand Ave @ 49th St 11/28/2000 NYCT10 Grand Ave @ 54th St 11/28/2000 NYCT11 Grand Ave @ 55th St 11/28/2000 NYCT12 Grand Ave @ Rust Ave 11/28/2000 NYCT13 Maurice Ave @ LIE (EB) off-ramp 5/21/1999 NYCDOS14 Maurice Ave @ LIE (WB) off-ramp 5/21/1999 NYCDOS15 Maurice Ave @ 58th St NYSDOT 516 Rust Ave @ Maspeth Ave 11/28/2000 NYCT17 48th St @ 56th Rd 5/21/1999 NYCDOS18 Review Ave/56th Rd @ Laurel Hill Blvd 5/21/1999 NYCDOS19 Meeker Ave south side @ Vandervoort Ave NYSDOT 520 Meeker Ave north side @ Vandervooft Ave NYSDOT 521 Greenpoint Ave @ Kingsland Ave 5/21/1999 NYCDOS

10/13/199922 Review Ave @ Greenpoint Ave 5/21/1999 NYCDOS

10/13/199923 48th St @ LIE EB off-ramp 5/21/1999 NYCDOS24 Laurel Hill Blvd/BQE @ 58th St NYSDOT 525 58th St @ Borden Ave/LIE service road 5/21/1999 NYCDOS26 55th Ave @ 58th St 5/21/1999 NYCDOS27 Review Ave @ Dept of Sanitation Site 5/21/1999 NYCDOS28 McGuiness Blvd @ Green Street 10/13/1999 NYCDOT 829 Jackson Ave @ 11th St. 10/13/1999 NYCDOT 830 11th St. @ 49th Ave 10/13/1999 NYCDOT 831 49th Ave @ 27th St 10/13/1999 NYCDOT 832 Metropolitan Ave @ Stewart Ave 10/13/1999 NYCDOT 833 Flushing Ave @ Williamsburg St West 3/25/1998 BNYDC

6/15/1999 NYCDCP 26/16/19996/17/1999

34 Flushing Ave @ Classon Ave 3/25/1998 BNYDC6/15/1999 NYCDCP 26/16/19996/17/1999

35 Kent Ave @ Williamsburg St West 3/25/1998 BNYDC36 Kent Ave @ Williamsburg St East 3/25/1998 BNYDC37 Kent Ave @ Classon Ave/Rutledge St 3/25/1998 BNYDC

1 See Table 1 for description of project data sources.Note: See Figure 3 for Available Turning Movement Counts

TABLE 4: AVAILABLE CLASSIFICATION COUNTSID Location Date of Count Data Source1

1 Manhattan Ave from Norman Ave to Meserole Ave 9/1998 NYSDOT 25/1999

10/20002 Borinquen Place Over Route 278 4/1998 NYSDOT 2

5/19998/2000

3 Metropolitan Ave Bridge Over English Kills 3/1998 NYSDOT 24 Greenpoint Ave from McGuinness Blvd to Kings/Queens Co Line 8/1998 NYSDOT 2

1 See Table 1 for description of project data sources.Note: See Figure 4 for Available Classification Counts

TABLE 5: PROPOSED ATR COUNT LOCATIONSID Roadway / Expressway Direction Location DescriptionA1 BQE EB Mainline before exit to Flushing AveA2 BQE EB Exit to Flushing AveA3 BQE EB Ent. from Williamsburg St E.A4 BQE EB Exit to Metropolitan AveA5 BQE EB Ent. from Williamsburg BridgeA6 BQE EB Exit to Humboldt StA7 BQE EB Exit to McGuinness BlvdA8 BQE EB Ent. from Meeker Ave/Vandervoort AveA9 BQE EB Mainline after Meeker Ave ent./Vandervoort AveA10 BQE EB Exit to LIE service roadA11 LIE EB Exit to EB BQEA12 BQE EB Exit to WB LIEA13 BQE EB Exit to 48th StA14 LIE service road EB Ent. from 48th StA15 LIE service road EB Exit to Maurice AveA16 LIE EB Ent. from Borden Ave/60th StA17 LIE EB Mainline between 48th St. and 58th St.A18 LIE WB Mainline between 48th St. and 58th St.A19 LIE WB Exit to Maurice Ave/63rd StA20 LIE WB Exit to service road for BQEA21 LIE service road WB Ent. from Borden AveA22 LIE service road WB Exit to LIE WB mainlineA23 LIE service road WB Exit to 48th StA24 LIE service road WB Exit to BQE WBA25 BQE EB Mainline after ent. from EB LIEA26 BQE WB Mainline after ent. from 61st StA27 BQE WB Exit to WB LIE/Hunters Point AveA28 LIE WB Exit to Greenpoint Ave/39th StA29 LIE WB Ent. from Van Dam StA30 LIE WB Exit to Van Dam StA31 LIE EB Exit to WB BQEA32 BQE WB Ent. from 51st Ave/43rd StA33 BQE WB Mainline after Kosciuszko BridgeA34 BQE WB Exit to Meeker AveA35 BQE WB Ent. from Meeker Ave/McGuinness BlvdA36 BQE WB Ent. from McGuinness BlvdA37 BQE WB Exit to Metropolitan AveA38 BQE WB Ent. from Metropolitan Ave/Marcy AveA39 BQE WB Exit to Williamsburg BridgeA40 BQE WB Exit to Flushing Ave/Kent Ave/Williamsburg StA41 BQE WB Ent. from Flushing AveA42 BQE WB Mainline after ent. from Flushing AveA43 Kent Ave NB/SB Between S11th St & S8th StA44 Grand St EB/WB Between Bushwick Ave & Waterbury StA45 Metropolitan Ave EB/WB Between Bushwick Ave & Olive StA46 Morgan Ave NB/SB Between Richardson St & Maspeth AveA47 Vandervoort Ave NB/SB Between Richardson St & Maspeth AveA48 Grand Ave EB/WB Between Gardner Ave & 47th StA49 Flushing Ave EB/WB Between Metropolitan Ave & Woodward AveA50 Page Pl NB/SB Between Grand Ave & Maspeth AveA51 Rust Ave NB/SB Between Grand Ave & 58th RdA52 Flushing Ave EB/WB Between Grand Ave & 59th StA53 Kent Ave NB/SB Between N10th St & N9th St

TABLE 5: PROPOSED ATR COUNT LOCATIONS (CONTINUED)ID Roadway / Expressway Direction Location DescriptionA54 McGuinness Blvd NB/SB Between Norman Ave & Nassau AveA55 McGuinness Blvd NB/SB Between Huron St & India StA56 McGuinness Blvd NB/SB Between Pulaski Bridge & Freeman StA57 Greenpoint Ave EB/WB Between Franklin St & Manhattan AveA58 Greenpoint Ave EB/WB Between Humboldt St & Russel AveA59 Jackson Ave EB/WB Between 21st St & 11th StA60 Van Dam St NB/SB Between 47th Ave & 48th AveA61 Greenpoint Ave EB/WB Between Bradley Ave & LIE EB service roadA62 Queens Blvd EB Between 58th St & 59th StA63 Queens Blvd WB Between 58th St & 59th StA64 Maurice Ave NB/SB Between 55th Dr. & Borden AveA65 LIE EB Exit to service road to WB BQE A66 Borden Ave EB/WB Between 23rd St & 21st StA67 Flushing Ave EB/WB Between Norstand Ave & Union AveA68 LIE EB Mainline after exit to 21st StA69 LIE WB Mainline after ent. from Van Dam StA70 LIE WB Mainline HOV lane after ent. from Van Dam StA71 Meeker Ave EB Between McGuinnes Blvd & Porter AveA72 Meeker Ave WB Between McGuinnes Blvd & Porter AveA73 Meeker Ave EB Between Metropolitan Ave & McGuinnes Blvd A74 Meeker Ave WB Between Metropolitan Ave & McGuinnes Blvd

Note: See Figure 5 for Proposed ATR Count Locations

TABLE 6: PROPOSED TURNING MOVEMENT COUNT LOCATIONSID IntersectionT1 Kent Ave @ Williamsburg St WT2 Kent Ave @ Willamsburg St ET3 Metropolitan Ave @ Marcy AveT4 Meeker Ave WB @ Metropolitan AveT5 Meeker Ave EB @ Metropolitan AveT6 Metropolitan Ave @ Vandervoort AveT7 Metropolitan Ave @ Stewart AveT8 Meeker Ave EB @ Humboldt StT9 Meeker Ave WB @ McGuinness BlvdT10 Meeker Ave WB @ Kingsland AveT11 Meeker Ave EB @ Kingsland AveT12 Meeker Ave WB @ Morgan AveT13 Meeker Ave EB @ Morgan AveT14 Meeker Ave WB @ Vandervoort AveT15 Meeker Ave EB @ Vandervoort AveT16 Greenpoint Ave @ Franklin StT17 Greenpoint Ave @ Manhattan AveT18 Greenpoint Ave @ McGuinness BlvdT19 Greenpoint Ave @ Humboldt AveT20 Greenpoint Ave @ Kingsland AveT21 Greenpoint Ave @ Review Ave/Van Dam StT22 Greenpoint Ave @ LIE EB service roadT23 Greenpoint Ave @ LIE WB service roadT24 Greenpoint Ave @ Hunters Point AveT25 Greenpoint Ave @ Queens BlvdT26 Franklin St @ India StT27 Jackson Ave @ Hunters Point Ave/11th StT28 Jackson Ave @ Thomson AveT29 Queens Blvd @ 69th StT30 Grand Ave @ 69th StT31 Grand Ave @ LIE WB service roadT32 Grand Ave @ LIE EB service roadT33 Maurice Ave @ LIE WB service roadT34 Maurice Ave @ LIE EB service roadT35 Maurice Ave @ Maspeth Ave/58th StT36 Grand Ave @ Page PlT37 Metropolitan Ave @ Flushing AveT38 Grand Ave @ Stewart AveT39 Metropolitan Ave @ Gardner AveT40 Metropolitan Ave @ Morgan AveT41 Grand St @ Morgan AveT42 Metropolitan Ave @ Humboldt AveT43 Grand St @ Bushwick StT44 Flushing Ave @ Humboldt StT45 Grand Ave @ Rust StT46 58th St @ LIE service road EB & WB/Borden AveT47 Van Dam St @ Hunters Point AveT48 Van Dam St @ LIE WB service roadT49 Van Dam St @ LIE EB service roadT50 Grand St @ Vandervoort AveT51 Flushing Ave @ Classon Ave/Rutledge StT52 McGuinness Blvd @ India StT53 Maurice Ave @ 69th St

TABLE 6: PROPOSED TURNING MOVEMENT COUNT LOCATIONS (CONTINUED)ID IntersectionT54 Borden Ave @ Queens Midtown Tunnel ent./exitT55 McGuinness Blvd @ Freeman StT56 48th St @ Laurel Hill Blvd EB & WBT57 48th St @ Exit Ramp from EB BQE

Note: See Figure 6 for Proposed Turning Movement Count Locations

TABLE 7: PROPOSED VEHICLE CLASSIFICATION AND OCCUPANCY COUNT LOCATIONSID Location DescriptionO1 Meeker Ave On-Ramp to BQE EB @ Vandervoort Ave (A8)O2 BQE EB after Meeker Ave entrance (A9)O3 EB BQE Off-Ramp to LIE WB (A12)O4 EB BQE Off-Ramp to LIE EB (A13)O5 LIE WB service road exit to WB BQE (A24)O6 LIE EB service road exit to WB BQE (A31)O7 BQE WB entrance from Borden Ave/43rd St (A32)O8 BQE WB before Meeker Ave exit (A33)O9 Meeker Ave Off-Ramp from BQE WB @ Apollo St (A34)

Note: See Figure 7 for Proposed Vehicle Classification and Occupancy Count Locations

TABLE 8: PROPOSED TRAVEL TIME AND DELAY RUN ROUTES1. Brooklyn-Queens Expressway (BQE)

- EB from entrance @ Tillary Street to exit to Queens Blvd- WB from exit to Queens Blvd WB to exit to Tillary St/Manhattan Bridge

2. Long Island Expressway (LIE)- EB from Queens Midtown Tunnel entrance @ plaza to LIE Exit to Queens Blvd- WB from Queens Blvd entrance to exit to Van Dam St

3. BQE & LIE connections- BQE EB from Vandervoort Ave entrance to entrance to EB LIE via LIE EB service road- LIE WB exit to WB BQE via LIE WB service road to LIE entrance to WB BQE

4. Queens Midtown Tunnel (QMT)- Review Ave from 37th St. - Van Dam St entrance to LIE WB - QMT - first exit in Manhattan- QMT/LIE EB entrance from Manhattan -LIE EB Exit to 48th St. - Laurel Hill

5. Meeker Ave- EB from Metropolitan Ave to entrance to BQE after Vandervoort Ave- WB from BQE exit to Meeker Ave to Metropolitan Ave

6. Metropolitan Ave- EB from Kent Ave to Vandervoort Ave- WB from Vandervoort Ave to Kent Ave

7. Grand St/Grand Ave- EB from Marcy Ave to LIE- WB from LIE to Marcy Ave

8. Flushing Ave- EB from Classon Ave/Rutledge St to Grand Ave- WB from Grand Ave to Classon Ave/Rutledge St

9. Greenpoint Ave- EB from Kent Ave to Queens Blvd- WB from Queens Blvd to Kent Ave

10. Kent Ave- NB from Flushing Ave to Green St- SB from Green St to Flushing Ave

11. McGuinness Blvd/Pulaski Bridge- NB from BQE/Meeker Ave EB to Jackson Ave- SB from Jackson Ave to BQE/Meeker Ave EB

12. Humboldt St & Kingsland Ave- Kingsland Ave NB from Maspeth Ave to Greenpoint Ave- Humboldt St SB from Greenpoint Ave to Flushing Ave

13. Vandervoort Ave- NB from Grand St to Meeker Ave/BQE WB- SB from Meeker Ave/BQE EB to Grand St

14. Rust St/56th Road/Review Ave- EB from Greenpoint Ave to Grand Ave- WB from Grand Ave to Greenpoint Ave

15. Maurice Ave/Page Pl- EB from Grand Ave to 69th St- WB from 69th St to Grand Ave

Note: See Figure 8 for Proposed Travel Time and Delay Run Routes

TABLE 9: PROPOSED PEDESTRIAN CROSSWALK COUNT LOCATIONSID IntersectionP1 Meeker Ave WB @ Metropolitan Ave (T4)P2 Meeker Ave EB @ Metropolitan Ave (T5)P3 Meeker Ave EB @ Humboldt St (T8)P4 Meeker Ave WB @ McGuinness Blvd (T9)P5 Greenpoint Ave @ McGuinness Blvd (T18)P6 Greenpoint Ave @ Queens Blvd (T25)P7 Grand Ave @ LIE WB service road (T31)P8 Grand Ave @ LIE EB service road (T32)P9 Maurice Ave @ LIE WB service road (T33)P10 Maurice Ave @ LIE EB service road (T34)

Note: See Figure 9 for Proposed Pedestrian Crosswalk Count Locations

EAST RIVER

NEWTOW N CREE

K

KOS

CIU

SZKO

BRID

GE

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BUR

GBR

IDG

E

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE .

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

Figu

re 1

Kos

cuis

zko

Bri

dge

Proj

ect T

raff

ic S

tudy

Are

aLEGEND

Prim

ary

Stu

dy A

rea

Seco

ndar

y St

udy

Area

Riv

er/C

reek

04,

000

2,00

0

Feet

1 in

ch e

qual

s 4,

000

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

# *

# *# *

# *

# *

# *

# *

# *

# *

# *# *

# *# *

# *

# *# *

# *

# *

# *

# *

# *

# *# *

# *

# *

# *

# *# *

# *# *

# *# *

# *

# *

# *

# *

# *

# *

# *

# *# *

EAST RIVER

NEWTOWN CREEK KO

SCIU

SZK

OBR

IDG

E

MA

NH

ATT

AN

BRID

GE

WIL

LIA

MS

BU

RG

BRID

GE

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UE

ENS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

POLI

TAN

AVE

.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

3

1

9

8

7

6

5

4

2

3433

41

40

39

38

37

36 35

3231

30

29

28

27

2625

24

23

22

20

1817

16

15

14

13

12

11

10

2119

Figu

re 2

Ava

ilabl

e A

utom

atic

Tra

ffic

Rec

orde

r (A

TR

)C

ount

Loc

atio

ns

LEGEND

# *Au

tom

atic

Tra

ffic

Rec

orde

r

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

See

Tabl

e 2

for s

peci

fic s

ourc

e an

d lo

catio

n in

form

atio

n.

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

$ + $ +$ +

$ +$ +

$ +$ +

$ +$ +

$ +$ +

$ +

$ +$ +

$ +

$ +

$ +$ +

$ +

$ +

$ +

$ +

$ + $ +

$ +

$ + $ +

$ +

$ +

$ +$ +$ + $ +$ +

$ +

$ +

EAST RIVER

NEWTOWN CREEK KO

SC

IUSZ

KOBR

IDG

E

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BUR

GBR

IDG

E

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

98

7

64

32

1

17

15

3736

35

3433

32

3130

29 28

2625

24

23

22

21

20

19

18

16

14

13

1211

510

Figu

re 3

Ava

ilabl

e T

urni

ng M

ovem

ent C

ount

(TM

C) L

ocat

ions

LEGEND

$ +Tu

rnin

g M

ovem

ent C

ount

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

See

Tabl

e 3

for s

peci

fic s

ourc

e an

d lo

catio

n in

form

atio

n.

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

!(

!(

!(

!(

EAST RIVER

NEWTOWN CREEK KO

SCIU

SZK

OBR

IDG

E

MA

NH

ATT

AN

BRID

GE

WIL

LIA

MS

BU

RG

BRID

GE

FLU

SH

ING

AV

E.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UE

ENS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

POLI

TAN

AV

E.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD 1

4

3

2

Figu

re 4

Ava

ilabl

e C

lass

ifica

tion

Cou

nt L

ocat

ions

LEGEND

!(C

lass

ifica

tion

Cou

nt

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

Base

data

cop

yrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub-

licen

sed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

See

Tabl

e 4

for s

peci

fic s

ourc

e an

d lo

catio

n in

form

atio

n.

# *# *# *

# *

# *# *# *

# *# *

# *# *

# *# *# *

# *# * # * # *# *

# *

# * # *

# * # *

# *# *

# *# *

# * # *# *

# *

# *# *

# *

# *

# *# *

# *

# *

# *# *

# *

# *

# *# *

# *# *

# *# *

# * # *# *

# *# * # *

# *# *# *# *# *

# *# *

# *# *

# *# *

# *

# *# *

# *# *# *

# *

EAST RIVER

NEWTOWN CRE EK

KOS

CIU

SZK

OBR

IDG

E

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BU

RG

BRID

GE

FLU

SH

ING

AV

E.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE

.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

A9A8

A6A7

A3

A4

A5

A2

A1A42

A19

A20

A16

A21

A15 A64

A17

A18

A63

A62

A25

A26

A23

A24 A22

A11 A1

2 A13

A10

A14

A27

A32

A31A28

A65

A60

A61

A30

A29

A69

A68A70

A66

A59

A58

A57

A52

A51

A50

A48

A49

A44

A45

A47

A46

A34

A71

A72

A54

A35

A53

A74 A73

A37

A40

A38

A39

A43

A67

A41

A56

A55

A33

A36

Figu

re 5

Prop

osed

Aut

omat

ic T

raff

ic R

ecor

der

(AT

R)

Cou

nt L

ocat

ions

LEGEND

# *Au

tom

atic

Tra

ffic

Rec

orde

r

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

See

Tabl

e 5

for s

peci

fic o

catio

n in

form

atio

n.

Base

data

cop

yrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub-

licen

sed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

%,

%,

%, %,

%,

%,

%,

%,

%, %,

%, %,

%, %,

%,

%,%,

%,

%,

%, %,%,%,

%, %,%,

%,

%,%,

%,

%,%,

%,%,

%,

%,

%,

%,%,

%,

%, %,

%, %,

%,%,

%,

%,

%,

%,

%,

%,

%,

%,

%,

%,

%,EAST RIVER

NEWTOWN CREEK KO

SC

IUSZ

KOBR

IDG

E

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BUR

GBR

IDG

E

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

T9

T3T5T4

T8

T6

T7

T2

T1

T56

T57

T34

T36

T44

T51

T35

T27

T15

T14

T11

T10

T18

T17

T54

T21

T20

T48

T49T22

T23

T25

T29

T53

T33

T32T31

T30

T50

T38

T43

T42

T39

T45

T46

T13

T12

T41

T40

T24

T47

T19

T16

T26

T28

T37

T52

T55

Figu

re 6

Prop

osed

Tur

ning

Mov

emen

t Cou

nt (T

MC

) Loc

atio

nsLEGEND

%,Tu

rnin

g M

ovem

ent C

ount

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

See

Tabl

e 6

for s

peci

fic o

catio

n in

form

atio

n.

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

! (

! (

! (! (

! (! ( ! (

! (

! (

EAST RIVER

NEWTOWN CREEK

KOS

CIU

SZKO

BRID

GE

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BUR

GBR

IDG

E

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

O9

O7

O6

O5 O4

O3

O2

O1

O8

Figu

re 7

Prop

osed

Veh

icle

Cla

ssifi

catio

n an

d O

ccup

ancy

Cou

nt L

ocat

ions

LEGEND

! (Ve

hicl

e C

lass

ifica

tion

and

Occ

upan

cy C

ount

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

See

Tabl

e 7

for s

peci

fic o

catio

n in

form

atio

n.

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

EAST RIVER

NEWTOWN CREEK

KOS

CIU

SZKO

BRID

GE

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BUR

GBR

IDG

E

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

" )7

" )7

" )8

" )8

" )10

" )10

" )9

" )9" )11

" )11

" )12

" )12" )6

" )6

" )14

" )14

" )15

" )15

" )1

" )1

" )3

" )4

" )4

" )4

" )5

" )5" )13

" )13

" )2

" )2

" )3

! (8

! (4

! (2! (4

! (4

! (4! (2

! (6! (10

! (9! (12

! (12

! (14

! (14

! (13

! (7

! (15

! (1! (11

! (12

! (9

! (5

! (3! (1

! (3

Figu

re 8

Prop

osed

Tra

vel T

ime

and

Del

ay R

un R

oute

sLEGEND

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

1 B

QE

2 LI

E3

BQ

E &

LIE

Con

nect

ions

4 Q

MT

5 M

eeke

r Ave

6 M

etro

polit

an A

ve7

Gra

nd S

t/Gra

nd A

ve8

Flus

hing

Ave

9 G

reen

poin

t Ave

10 K

ent A

ve11

McG

uine

ss A

ve/P

ulas

ki B

ridge

12 H

umbo

ldt A

ve &

Kin

gsla

nd A

ve13

Van

derv

oort

Ave

14 R

ust S

t/56t

h/R

evie

w A

ve15

Mau

rice

Ave

Riv

er/C

reek

" )1 ! (1En

dpoi

nt R

oute

Mar

ker

Inte

rmed

iate

Rou

te M

arke

r

See

Tabl

e 8

for m

ore

deta

il on

rout

es.

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

! A! A

! A

! A! A

! A

! A

! A! A

! A

EAST RIVER

NEWTOWN CREEK KO

SC

IUSZ

KOBR

IDG

E

MA

NH

ATTA

NBR

IDG

E

WIL

LIA

MS

BUR

GBR

IDG

E

FLU

SH

ING

AVE

.

GR

AND

ST

BQE

QU

EEN

S B

LVD

.Q

UEE

NS

-MID

TOW

NTU

NN

EL

LIE

ME

TRO

PO

LITA

N A

VE.

GR

AND

AVE

BQE

GRE

ENPO

INT

AVE

MCGUINESS BLVD

P9

P8

P7

P6

P5

P4P3

P1P2

P10

Figu

re 9

Pede

stri

an C

ross

wal

k C

ount

Loc

atio

nsLEGEND

! APe

dest

rian

Cro

ssw

alk

Cou

nt

Riv

er/C

reek

03,

700

1,85

0

Feet

1 in

ch e

qual

s 3,

700

feet

ZNe

w Yo

rk S

tate

Depa

rtmen

t of T

rans

porta

tion

See

Tabl

e 9

for s

peci

fic o

catio

n in

form

atio

n.

Bas

edat

a co

pyrig

hted

by

City

of N

ew Y

ork

Dep

artm

ent o

f Env

irnm

enta

l Pro

tect

ion,

Sub

-lice

nsed

from

New

Yor

k St

ate

Dep

artm

ent o

f Tra

nspo

rtatio

n - R

egio

n 11

.

TECHNICAL MEMORANDUM T3:

TRAVEL DEMAND/TRAFFIC SIMULATION

MODELING METHODOLOGY

Performed for the New York State Department of Transportation In Support of the

Draft Environmental Impact Statement For the Kosciuszko Bridge Project

Kings County and Queens County, NY

September 2006

DRAFT

Kosciuszko Bridge Project

New York State Department of Transportation

Kosciuszko Bridge Project 1 September 2006

A. INTRODUCTION

The Federal Highway Administration (FHWA), in cooperation with the New York State Department of Transportation (NYSDOT), is preparing an Environmental Impact Statement (EIS) for the Kosciuszko Bridge Project in Brooklyn/Queens. The purpose of this project is to identify the structural, safety, and operational deficiencies of the bridge, and to provide a cost-effective Newtown Creek crossing capable of accommodating future traffic demand, as well as supporting the operational efficiency of the Kosciuszko Bridge corridor.

Integral to the project was the development and application of an integrated travel demand/traffic simulation modeling system to be used as the means to predict and evaluate how well future alternative improvements achieve the project goals, such as providing safe and efficient transportation services at a reasonable cost with minimal environmental impact. More specifically, the modeling system was employed to:

Identify existing and future year traffic operational deficiencies in the study area;

Assess alternative improvement proposals on the Kosciuszko Bridge corridor; and

Provide performance measures for use in environmental analyses.

Since the essential role of the model is to help decision-makers determine the preferred improvements/programs and set priorities to address anticipated traffic problems, failure to develop a reliable model can lead to erroneous decisions for future transportation investments, resulting in the inefficient use increasingly scarce transportation funds. Recognizing this, the purpose of this Technical Memorandum is two-fold: (1) to outline an acceptable methodology for establishing traffic forecasting and operational analysis procedures, and (2) to document the model calibration process and the results obtained. The proposed methodology and model calibration build on currently available state-of-the-art modeling techniques, with the goal of providing reliable traffic forecasting to meet the Federal mandates and standards for accuracy.

The remainder of this Technical Memorandum is organized as follows. It begins with a brief description of the general structure of the model system as well as basic preparations for model development. Next, it depicts the methodologies and assumptions used to establish a base year (2002) model system that would replicate existing travel patterns and traffic operations in the Kosciuszko Bridge corridor. This is followed by a discussion of the calibration process and results for both travel demand and traffic simulation models. Finally, conclusions are drawn.

B. MODEL STRUCTURE

To meet the project needs, a set of mathematical models and computer programs were developed to predict how travel patterns/traffic operations in the project study area would change in response to future changes in land use patterns, demographics, and the region’s transportation facilities and services (including the Kosciuszko Bridge design alternatives). As shown in Figure 1, the general model system structure is composed of three transportation models linked by interface programs facilitating the transfer of data between models. These three models include the BPM regional travel demand model, BPM sub-area network model, and VISSIM traffic simulation model. Each model is briefly described below.

Travel Demand/Traffic Simulation Modeling Methodology


BPM/TransCAD

MacroscopicRegional

Travel DemandModel

BPM/TransCADMacroscopic Sub-area

Network Model

VISSIMMicroscopic Sub-area

Simulation Model

DataConversion

DataConversion

FIGURE 1: MODEL STRUCTURE

BPM Regional Travel Demand Model – used to predict how region-wide transportation system changes may affect individuals’ travel patterns including when and where to travel, which mode to use, and which roadways or transit lines to use. The end-result of the model is the travel pattern predictions (by time of day) regarding traffic volumes, transit ridership and other data such as average travel time and vehicle-miles traveled. For this project, the regional travel demand modeling process was based on the Best Practice Model (BPM) developed by the New York Metropolitan Transportation Council (NYMTC).

Based on the concept of “tours” rather than traditional “trips”, the BPM can model a full range of travel decisions, including auto ownership, journey frequency, primary destination, journey stop, mode choice, time of day, and route choice. The model area covers 28 counties in the tri-state region: New York, New Jersey, and Connecticut. This regional coverage would sufficiently internalize the New York City metropolitan area to reflect the current labor market. The BPM is composed of 3,500 transportation analysis zones and includes all minor arterials and above, and all forms of public transportation.

BPM Sub-Area Network Model – used to provide more accurate forecasts of roadway traffic volumes and/or transit line volumes at a corridor or sub-area level. The rationale of a sub-area model is based on the observation that traffic impact usually diminishes away from a project site (i.e., the Kosciuszko Bridge). In addition, sub-area modeling has the following advantages. First, supply and demand variables can be represented with more detail. For example, additional detail can be added to the network within the sub-area. Second, relationships in the models can be calibrated more precisely. Finally, tremendous savings in computer time can be achieved during the model calibration.

For this project, the sub-area models are essentially the highway assignment models and hence the other BPM model components are fixed inputs for these models. A sub-area network was extracted from the BPM regional network, with external stations established to provide the linkages between the sub-area and the rest of the region. The sub-area trip tables were obtained through the use of “Sub-area Analysis” in TransCAD. Notice that for each design alternative, the BPM regional model would be



run first and the sub-area network and trip tables were then extracted from the regional model. Hence, it is possible to analyze sub-area traffic patterns in detail while the overall effects of design alternatives on the regional transportation system are also captured.

VISSIM Sub-Area Traffic simulation model – used to determine how well the roadway system would operate under alternative transportation improvements, including geometric designs (e.g., ramp configuration), control strategies (e.g., signal timing), and operating policies (e.g., HOV lanes). Notice that the traffic flow predictions from travel demand models only offer a static average view of the expected use of the improvement projects, and these non-time-varying traffic flows can not accurately represent oversaturated conditions on the roadways during the peak periods. Hence a traffic simulation model is used to complement the travel demand model by translating the changes in travel patterns into more meaningful measures of travel impacts on motorists, such as level of service, total delay, and average speed, etc.

For this project, the micro-simulation model VISSIM was used because it can analyze the full range of functionally classified roadways and transit operations and provide more realistic simulation of vehicle movements on the network (due to VISSIM’s unique features of link-connector based network and psycho-physical driver behavior logic). In addition, VISSIM’s 3-D animation capability of displaying the simulated traffic operations is highly effective to convey the advantages and disadvantages of the project alternatives to both technical and non-technical audiences.

In summary, the BPM regional travel demand model serves as the starting point for estimating changes in regional travel patterns that would result from transportation system changes or economic growth changes for the region. To generate accurate forecasts of vehicular traffic volumes in the Kosciuszko Bridge corridor, a sub-area model is developed by extracting the sub-area networks and trip tables from the BPM regional model. The sub-area highway assignment results (including link volumes, turning movements, traffic compositions, and fixed route information) are then used to establish traffic demands and route choice for the use of VISSIM simulation. The VISSIM model provides more detailed forecasts of the roadway performance, such as intersection level-of-service, travel time and delay on highway segments, etc.

C. MODEL PREPARATION

Preparations for model development and forecasting include identification of key input data for model development and calibration, and determination of study areas, analysis years (base/future years), and analysis periods (AM/PM peak hours) for traffic analysis and forecasting. These various preparations are described below.

C.1. Study Area

The project study area was divided into primary, secondary, and tertiary study areas to differentiate the modeling level of detail on their respective network linkages. The primary study area was defined as the area where a major shift in local traffic using alternate routes would occur as a result of the rehabilitation or replacement of the Kosciuszko Bridge. The primary study area was roughly bounded by the Long Island Expressway (LIE), Grand Street/Grand Avenue, and the East River.



The secondary study area was defined as the area where a potential shift in mode choice and/or route choice would occur as a result of improvements to the Kosciuszko Bridge. The geographic scope and level of detail were developed to permit the evaluation of a wide range of multimodal and intermodal transportation strategies and their resulting impacts on the Kosciuszko Bridge travel demand. The secondary study area was bounded by Skillman Avenue/Woodside Avenue, Flushing Avenue, and the East River.

The tertiary study area was defined to have a region-wide geographic coverage encompassing the long-distance travel movements, which affect or are affected by the Kosciuszko Bridge travel conditions. For simplicity, the tertiary study area was assumed to be identical to the BPM regional model area (that covers 28 counties in the tri-state region). This wide-area geographic coverage would allow the investigation of the regional travel choices available for diverting auto and truck traffic to or from the Kosciuszko Bridge. Since the tertiary study area is extremely large, Figure 2 illustrates the primary and secondary study areas only.

FIGURE 2: PRIMARY AND SECONDARY STUDY AREAS

C.2. Base Year/Future Year Determination

The base year of the model analysis is 2002, corresponding to the year in which the traffic data in the study area were collected. The future years selected for model applications include the estimated time of completion (ETC) (2015), ETC+10(2025), ETC+20(2035), and ETC+30 (2045), the project’s design year. The design year is the horizon year specified and used by engineers to represent the end of the economic life of a proposed transportation improvement. Both ETC+10 (2025) and ETC+20 (2035) years were selected for the requirement of air quality analysis.



C.3. Analysis Period

Due to the constantly oversaturated traffic conditions in the Kosciuszko Bridge corridor, the overall peak hour level of service might remain unchanged for many hours in the morning and evening. Therefore, an extended analysis period was considered to be unnecessary for the purpose of discerning congestion build-up and dissipation. For this project, travel demand would be represented by two separate peak hours:

AM Peak Hour: 6:45 – 7:45 a.m.; and

PM Peak Hour: 4:45 – 5:45 p.m.

For traffic simulation, the analysis period is one (peak) hour, but simulation was performed for more than one hour to include the “warm-up period”. Empirical testing revealed that a 15-minute warm-up period would achieve model equilibrium.

C.4. Data Inventory and Collection

Data collection is required to provide the travel demand/traffic simulation model input parameters and output measures of performance for model calibration and validation. Since the BPM has prepared most of the input data for demand forecasting, it needs data mainly for model calibration and variable validation. For example, information on existing and future land use patterns (including the location and intensity of use) is required for validation of the projected socio-economic data contained in the BPM. Future committed/programmed highway and transit improvement projects were assembled for updating the transportation networks. For model calibration, traffic volume counts and vehicle classification counts were used to determine whether the base year model accurately reflects existing conditions (i.e., number and composition of vehicle types traveling in the study area).

Input data to the traffic simulation model can be grouped into supply, demand and control. Demand data represents varied demand profiles (by vehicle type) over the peak periods, and was generated by the BPM sub-area model, along with automatic traffic recorder (ATR) volume counts and field classification counts. Control data include stop/yield sign control and each signal’s phasing/timing plan. Supply data consist of design and traffic flow characteristics of each link and node, such as link length, number of lanes, lane width, grade, curvature, speed limit, lane restriction/prohibition, etc. Based on the BPM sub-area network, aerial photography, as-built plans, and field reconnaissance, a simulation network was developed with roadway layout and geometry consistent with the actual network. For model calibration, traffic volume and travel time data would be used to ensure that the base year model accurately reflects the existing conditions.

D. MODEL DEVELOPMENT (BASE YEAR)

This section describes the methodologies and assumptions used to establish a base year (2002) model system that would replicate the existing travel patterns and traffic operations in the Kosciuszko Bridge corridor. Before the model can be used to estimate future conditions, it must closely represent the existing conditions that it describes. To meet this requirement, an integrated travel demand/traffic operation platform was designed to provide an effective modeling framework for the base year model development. Figure 3 illustrates the major



elements of this framework. Various steps used for the base year model development are described as follows:

Socio-ecoData

Household SynthesisAuto Ownership

Journey Production

ModePrimary Destination

Secondary Stops

ExternalAutos/Trucks

Time-ofDay

Trip Assignment

Regional CalibrationAdaptive Assignment/CoreModel Adjustment Factors/

Assignment Parameters

Regional Travel Demand ModelBPM

RegionalNetworks

Sub-AreaNetworks/Trip Tables

Assignment Parameters/Look-up Table

Sub-Area ExtractionTransCAD

Sub-Area Networks/TripTables/Assignment

Parameters/Look-up Table

Assignment/Calibration

PathInformation

Pre-simulationProcessor

Sub-Area Network ModelBPM

Sub-Area Networks/TrafficFlows/Path Information/Priority Rules/Detectors

Simulation/Calibration

Sub-Area Simulation ModelVISSIM

BPM Sub-Area/VISSIM Interface

BPM Regional/Sub-Area Interface

PerformanceMeasures(MOEs)

FIGURE 3: BASE YEAR MODEL DEVELOPMENT PROCESS

Validate BPM regional highway/transit networks – The highway and transit networks were checked and refined to more precisely reflect the transportation systems in the project study area. In particular, the highway links and transit routes in the Kosciuszko Bridge corridor were reviewed extensively.

Validate BPM zonal socio-economic data – Travel demand (in terms of journey production and attraction) is greatly influenced by the socio-economic data, including population, employment, etc. Hence, these socio-economic variables were checked and validated at the zonal level, using information on the existing land use patterns (including the location and intensity of use) in the study area.

Run the BPM core (or regional) model – Application of the BPM for a full regional-scale model run was undertaken for the base year (2002) conditions. This full BPM model application includes the major runs of highway/transit skim and accessibility, household-auto-ownership-journey production (HAJ), mode-destination-stop choice (MDSC), external autos, commercial vehicles, time of day, and highway and transit assignments.

Calibrate and validate BPM regional models – To confirm whether the BPM results are in close agreement with actual year 2002 conditions, the assigned volumes were compared with the observed traffic counts at selected locations. Since the computer time for full a BPM model run is enormous, the traffic volume comparison at this stage



would focus on: (1) screenlines and/or cutlines established to intercept major traffic flows through the study area, and (2) external stations established on the boundary of the study area where significant amounts of trips enter and exit the study area. The BPM regional models were calibrated using the procedure of highway adaptive assignment and/or core model adjustment factors.

Develop BPM sub-area models – A sub-area network was extracted from the BPM regional network, with external stations established to provide the linkages between the sub-area and the rest of the region. The sub-area trip tables were obtained through the use of “Sub-area Analysis” in TransCAD. Sub-area modeling has the advantages that: (1) supply and demand variables can be represented with more detail; (2) relationships in the models can be calibrated more precisely, and (3) computer time for calibration is substantially less due to the smaller study area.

Calibrate and validate BPM sub-area models – Model calibration at this stage focused on all freeways and major arterials in the study area. The assigned base year trips were validated on a link-by-link basis and the models were refined until the percentage differences between assigned trips and actual trips were minimized.

Conduct Pre-Simulation Preparation in TransCAD – The BPM sub-area highway networks and their associated attributes were prepared for the purpose of developing VISSIM networks. The BPM sub-area highway assignment results (including link volumes, turning movements at intersections or junctions, traffic compositions, and fixed route information) were prepared to establish traffic demands and route choice for the use of VISSIM simulation.

Develop VISSIM model – Based on the BPM sub-area highway networks and aerial photography of the study area, VISSIM networks were established for both AM and PM peak hours. In addition to passenger vehicles, information on bus operations (including routes, stop locations, and headways/schedules) was also embedded into the simulation network. Various field data and sub-area assignment results were used to develop demand profiles throughout the simulation time period.

Run VISSIM simulation – Once the VISSIM model was developed, it was run for two scenarios, including AM and PM peak hours. For each scenario, the error checking procedure was undertaken by reviewing the on-screen animation and model outputs to determine the model’s accuracy in simulating field operations. The input coding error checking was also performed so that the later calibration process would not result in parameters that were distorted to compensate for overlooked coding errors. .

Calibrate and validate VISSIM model – Once the error checking and correction step were completed, the calibration process proceeded with a detailed numerical comparison of model results against field measurements, and appropriate fine-tuning of the various input parameters that describe traffic control operation, traffic flow characteristics, and driver behavior. The simulated trips in the networks were compared against the observed link volumes and travel speeds. The model calibration process was terminated when the discrepancy in volumes and speeds fell within an acceptable error.

Generate performance measures – Custom post-processing programs were developed to generate various performance measures (or measures of effectiveness



[MOEs]) from the BPM sub-area model and the VISSIM model. In the case of the VISSIM model, up to five simulation runs, using different random number seeds for each individual simulation run were conducted for each VISSIM application.

E. BPM SUB-AREA MODEL CALIBRATION

To ensure the BPM sub-area was model capable of predicting precise changes in traffic volumes that would result from transportation system changes, a model calibration was conducted to ensure that the base year sub-area modeling results would accurately replicate field measurements. Only highway assignment calibration for the sub-area model was undertaken with the goal of producing assigned volumes which matched, as closely as possible, the counted volumes for 2002 base year conditions. The purpose of this section is to describe the calibration procedures and results of the sub-area model.

E.1. Calibration Preparation

For the base year 2002, the sub-area (highway assignment) models were calibrated to represent the AM and PM peak hour traffic flows in the sub-area highway networks, respectively. Prior to model calibration, the following data inputs were prepared, including:

Highway network. The sub-area AM and PM peak hour highway networks (see Figure 4) and their associated link attributes were, respectively, transferred from the BPM AM and PM peak period highway networks with the exception of link capacity. The peak hour link capacity was calculated by multiplying the number of lanes by the hourly per-lane capacity. Also transferred from the BPM is a lookup table of free-flow speed and hourly capacity values classified by the physical link type and the area type of that link. This lookup table is used to differentiate between links of a particular facility type that may have different speed/capacity characteristics because of adjacent land use, access control, or design element differences.



FIGURE 4: BPM SUB-AREA NETWORK

Origin-destination (O-D) trip matrices. The sub-area AM and PM peak hour O-D trip matrices were developed, respectively, from the BPM AM and PM peak period trip O-D matrices using time-of-day factors derived from the 2002 ATR volume counts. Six O-D trip matrices were created, separately, for drive alone, HOV 2, HOV 3+, taxi, truck, and other commercial vehicle (e.g., van) trips. Bus trips were included in the HOV+3 O-D trip matrix.

Volume delay function. The volume delay function (VDF) in the BPM sub-area models used the Bureau of Public Roads (BPR) equation, which relates link speed as a function of the volume/capacity ratio. The parameter values of α and β in the BPR equation would vary by the physical link type, and were transferred from the BPM regional model directly.

Traffic counts. The observed AM and PM peak hour traffic volumes were obtained from ATR volume counts performed during a two-week period from November 11 to November 24, 2002.

Calibration criterion. The criterion for an acceptable level of calibration accuracy was established based on two documents: Calibrating and Adjustment of System Planning Models (FHWA-ED-90-015, Federal Highway Administration, December 1990) and Model Validation and Reasonableness Checking Manual (Transportation Model Improvement Program and Federal Highway Administration, February 1997).



E.2. Calibration Process

Model calibration is the process of adjusting the model parameters by comparing modeled results with observed data. In the case of trip assignment calibration, the model parameters adjusted include:

Free-flow speed (or free-flow travel time);

Capacity;

α and β values; and

O-D trip matrices.

Iterative adjustments were made to these parameters until the modeled traffic volumes were within the acceptable range of the observed ground counts. After each adjustment, a multi-modal multi-class assignment (MMA) procedure in TransCAD was used to load the six separate trip matrices onto the highway network simultaneously. A convergence value of 0.1 and a maximum of 20 iterations were specified for the MMA procedure.

The approach to the calibration process adopted a link-based validation analysis, a much higher standard for calibration than that used by the BPM regional model. Not only are overall flows of traffic volumes examined, but also link-specific volumes at critical locations. Due to the lack of balanced traffic data on local streets, the assignment calibration efforts were primarily focused on the major links of highways in the study area, including the BQE, and LIE mainlines and service roads. Based on the two documents mentioned above, the criteria for an acceptable level of calibration accuracy include:

A correlation coefficient (to be described below) would have a value greater than 0.88;

75% of all highway links should have the modeled link volumes within ±20% of the observed volumes;

50% of all highway links should have the modeled link volumes within ±10% of the observed volumes; and

The percent of root-mean-square error (to be described below) would have an appropriate aggregate value less than 30%.

E.3. Calibration Results and Statistical Validation

The results of the AM and PM peak hour model calibration are shown in Tables 1 and 2. These tables present a comparison of assigned traffic volumes to actual ground counts for all links in the highway network. To examine how closely assigned volumes matches observed volumes, correlation coefficients (r2) were first calculated using pairs of assigned and observed volumes as follows:



2

1

2

1

2

1

2

1

2

1 1 12

])(())([⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢

⎣

⎡

−−

−=

∑ ∑∑ ∑

∑ ∑ ∑

= == =

= = =

N

i

N

iii

N

i

N

iii

N

i

N

i

N

iiiii

OONSSN

OSOSNr

where

Si = assigned volume on link i,

Oi = observed volume on link i, and

N = total number of links with a count.

A correlation coefficient value lies between 0 and 1, and is desirable to be closer to one. As shown in Figures 5 and 6, both AM and PM peak hour models have a correlation coefficient of 0.99, which is substantially higher than the minimum value of 0.88 recommended by FHWA. As shown in Tables 1 and 2, the general correspondence between assigned volumes and observed volumes appears to be good except for some ramp links. More than 79% of the links with assigned volumes within ±10% of the counted volumes. The assignments to these high-volume links (i.e., larger than 3,000 vehicles) are quite accurate, with the assigned volumes within ±2.7% of the counted volumes (on the average) for AM peak hour and ±2.0% for PM peak hour. On an “actual difference” basis, the model results generally matched ground counts, with variations less than 200 vehicles (per hour). For these ramps with a relatively large percent difference, the majority have an actual difference lying only between 0 and 200 vehicles (per hour).



Figure 5: 2002 AM Peak HourTransCAD Volume vs Observed Volume

correlation coeff. = 0.9978

0

1000

2000

3000

4000

5000

6000

0 1000 2000 3000 4000 5000 6000

Observed Volume

Tran

sCA

D V

olum

e

`

FIGURE 5: 2002 A.M. PEAK HOUR TRANSCAD VOLUME VS. OBSERVED VOLUME

Figure 6: 2002 PM Peak HourTansCAD Volume vs Observed Volume


0

1000

2000

3000

4000

5000

6000

0 1000 2000 3000 4000 5000 6000

Observed Volume

Tran

sCAD

Vol

ume

FIGURE 6: 2002 P.M. PEAK HOUR TRANSCAD VOLUME VS. OBSERVED VOLUME

TABLE 1: SUB-AREA MODEL CALIBRATION RESULTS – AM PEAK HOUR

Link Name Counted Volume

Assigned Volume

Actual Difference

Percent Difference

Brooklyn-Queens Expressway - Eastbound

Mainline bet Tillary St entrance and Flushing Ave exit 3,770 3,621 -149 -3.95%

Exit ramp to Flushing Ave./Classon Ave. 580 557 -23 -3.97%

Mainline bet Flushing Ave exit & Williamsburg St entrance 3,190 3,064 -126 -3.95%

Entrance ramp from Williamsburg St. E 960 868 -92 -9.58%

Mainline bet Williamsburg St entrance & Rodney St exit 4,150 3,932 -218 -5.25%

Exit ramp to Rodney St 680 666 -14 -2.06%

Mainline bet Rodney St exit & Williamsburg Bridge entrance 3,470 3,266 -204 -5.88%

Entrance from Williamsburg Bridge 1,020 1,190 170 16.67%

Mainline bet Williamsburg Bridge entrance & McGuinness Blvd/Humboldt St exit

4,490 4,456 -34 -0.76%

Exit ramp to McGuinness Blvd/Humboldt St 1,140 1,082 -58 -5.09%

Exit ramp to McGuinness Blvd 960 984 24 2.50%

Exit ramp to Humboldt St 180 98 -82 -45.56%



Mainline bet Humboldt St exit & Meeker Ave entrance 3,350 3,374 24 0.72%

Entrance ramp from Meeker Ave 1,250 1,201 -49 -3.92%

Mainline bet Meeker Ave entrance & LIE exit 4,600 4,575 -25 -0.54%

Exit ramp to LIE WB mainline & LIE EB Service Rd (SR) 2,860 2,819 -41 -1.43%

Mainline bet LIE exit & LIE EB SR entrance 1,740 1,756 16 0.92%

Entrance ramp from LIE EB SR 520 552 32 6.15%

Mainline after LIE EB SR entrance 2,260 2,308 48 2.12%

Brooklyn-Queens Expressway – Westbound

Mainline before exit to LIE WB mainline/service road (SR) 3,380 3,226 -154 -4.56%

Exit to LIE WB mainline/SR 870 805 -65 -7.47%

Mainline bet LIE mainline/SR exit & LIE EB/WB SR/43rd St entrance

2,510 2,420 -90 -3.59%

Entrance from LIE EB/WB SR/43rd St 2,480 2,615 135 5.44%

Mainline bet LIE EB/WB SR/43rd St entrance & Vandervoort Ave exit

4,990 5,036 46 0.92%

Exit ramp to Vandervoort Ave 960 832 -128 -13.33%

Mainline bet Vandervoort Ave exit & McGuinness Blvd/Meeker Ave entrance

4,030 4,203 173 4.29%

Entrance from Meeker Ave 370 270 -100 -27.03%

Entrance from McGuinness Blvd 640 746 106 16.56%

Entrance ramp from McGuinness Blvd/Meeker Ave 1,010 1,017 7 0.69%

Mainline bet McGuinness Blvd/Meeker Ave entrance & Metropolitan Ave exit

5,040 5,220 180 3.57%

Exit ramp to Metropolitan Ave 410 525 115 28.05%

Mainline bet Metropolitan Ave exit & Williamsburg Bridge exit

4,630 4,694 64 1.38%

Mainline bet Williamsburg Bridge exit & Metropolitan Ave entrance

2,590 2,556 -34 -1.31%

Exit to Williamsburg Bridge 2,040 2,137 97 4.75%

Entrance ramp from Metropolitan Ave 510 521 11 2.16%

Mainline bet Metropolitan Ave entrance & Williamsburg St exit

3,100 3,078 -22 -0.71%

Exit ramp to Williamsburg St W 680 660 -20 -2.94%

Mainline bet Williamsburg St exit & Flushing Ave entrance 2,420 2,414 -6 -0.25%

Entrance ramp to Flushing Ave 610 498 -112 -18.36%

Mainline after Flushing Ave entrance 3,030 2,915 -115 -3.80%

Long Island Expressway – Westbound



Mainline before 65th St exit 6,390 6,166 -224 -3.51%

Exit ramp to 65th St 750 536 -214 -28.53%

Mainline bet 65th St exit & LIE WE SR exit 5,640 5,629 -11 -0.20%

Mainline bet LIE WB SR exit & mainline/HOV split 3,480 3,322 -158 -4.54%

Mainline bet mainline/HOV split & LIE WB SR exit 3,210 3,110 -100 -3.12%

Mainline bet LIE WB SR exit & BQE EB/WB entrance 1,540 1,433 -107 -6.95%

Entrance from BQE EB/LIE WB Service Rd 1,750 1,777 27 1.54%

Entrance from BQE WB 400 350 -50 -12.50%

Entrance from BQE EB/WB 2,150 2,128 -22 -1.02%

Mainline bet BQE EB/WB entrance & Gale Ave exit 3,690 3,561 -129 -3.50%

Exit ramp to Gale Ave 1,580 1,640 60 3.80%

Mainline bet Gale Ave exit & Van Dam St entrance 2,110 1,921 -189 -8.96%

Entrance ramp from Van Dam St 40 102 62 155.00%

Mainline after Van Dam St entrance 2,150 2,024 -126 -5.86%

Exit to LIE WB Service Rd 2,160 2,307 147 6.81%

Entrance ramp from Maurice Ave 1,440 1,242 -198 -13.75%

Service Rd bet Maurice Ave entrance & BQE WB/48th St exit 3,600 3,549 -51 -1.42%

Exit to BQE WB/48th St 2,080 2,102 22 1.06%

Exit ramp to 48th St 150 160 10 6.67%

Exit ramp to BQE WB 1,930 1,942 12 0.62%

Exit ramp to LIE WE mainline 1,520 1,446 -74 -4.87%

Entrance from LIE WB mainline 1,670 1,676 6 0.36%


Service Rd after entrance from BQE WB/LIE WB 2,140 2,131 -9 -0.42%

Long Island Expressway – Eastbound

Mainline before LIE EB Service Rd (SR) exit 1,540 1,540 0 0.00%

Exit ramp to LIE EB SR 400 437 37 9.25%

Mainline bet LIE EB SR exit & LIE EB SR entrance 1,140 1,103 -37 -3.25%

Entrance ramp from LIE EB SR 1,070 1,191 121 11.31%

Mainline bet LIE EB SR entrance & mainline/SR merge 2,210 2,294 84 3.80%

Mainline bet LIE EB mainline/SR merge & Maurice Ave entrance

4,460 4,385 -75 -1.68%

Entrance ramp from Maurice Ave 570 662 92 16.14%

Mainline after Maurice Ave entrance 5,030 5,047 17 0.34%



LIE EB SR before entrance ramp from LIE EB mainline 1,640 1,728 88 5.37%

SR bet LIE EB mainline entrance & BQE WB/ Laurel Hill Blvd exit

2,040 2,166 126 6.18%

Exit to BQE WB/ Laurel Hill Blvd 260 224 -36 -13.85%

Exit to Laurel Hill Blvd 50 66 16 32.00%

Exit to BQE WB 210 158 -52 -24.76%

SR bet BQE WB/ Laurel Hill Blvd exit & LIE EB mainline exit 1,780 1,941 161 9.04%

SR bet LIE EB mainline exit & BQE EB exit 710 750 40 5.63%

SR bet BQE EB exit & BQE EB entrance 190 197 7 3.68%

Entrance ramp from BQE EB 2,230 2,042 -188 -8.43%

SR bet BQE EB entrance & 48th St entrance 2,420 2,240 -180 -7.44%

Entrance ramp from 48th St 260 380 120 46.15%

SR bet 48th St entrance & Maurice Ave exit 2,680 2,621 -59 -2.20%

Exit ramp to Maurice Ave 430 530 100 23.26%

SR bet Maurice Ave exit & mainline/SR merge 2,250 2,090 -160 -7.11%

TABLE 2: SUB-AREA MODEL CALIBRATION RESULTS – PM PEAK HOUR

Link Name Counted Volume

Assigned Volume

Actual Difference

Percent Difference

Brooklyn-Queens Expressway - Eastbound

Mainline bet Tillary St entrance and Flushing Ave exit 3,730 3,780 50 1.34%

Exit ramp to Flushing Ave./Classon Ave. 790 840 50 6.33%

Mainline bet Flushing Ave exit & Williamsburg St entrance 2,940 2,940 0 0.00%

Entrance ramp from Williamsburg St. E 620 569 -51 -8.23%

Mainline bet Williamsburg St entrance & Rodney St exit 3,560 3,509 -51 -1.43%

Exit ramp to Rodney St 870 767 -103 -11.84%

Mainline bet Rodney St exit & Williamsburg Bridge entrance 2,690 2,741 51 1.90%

Entrance from Williamsburg Bridge 1,480 1,474 -6 -0.41%

Mainline bet Williamsburg Bridge entrance & McGuinness Blvd/Humboldt St exit

4,170 4,216 46 1.10%

Exit ramp to McGuinness Blvd/Humboldt St 820 856 36 4.39%

Exit ramp to McGuinness Blvd 630 692 62 9.84%

Exit ramp to Humboldt St 190 163 -27 -14.21%

Mainline bet Humboldt St exit & Meeker Ave entrance 3,350 3,360 10 0.30%

Entrance ramp from Meeker Ave 1,480 1,552 72 4.86%

Mainline bet Meeker Ave entrance & LIE exit 4,830 4,912 82 1.70%



Exit ramp to LIE WB mainline & LIE EB Service Rd (SR) 2,540 2,586 46 1.81%

Mainline bet LIE exit & LIE EB SR entrance 2,290 2,325 35 1.53%

Entrance ramp from LIE EB SR 300 364 64 21.33%

Mainline after LIE EB SR entrance 2,590 2,690 100 3.86%


Mainline before exit to LIE WB mainline/service road (SR) 2,760 2,753 -7 -0.25%

Exit to LIE WB mainline/SR 680 648 -32 -4.71%

Mainline bet LIE mainline/SR exit & LIE EB/WB SR/43rd St entrance

2,080 2,105 25 1.20%

Entrance from LIE EB/WB SR/43rd St 2,650 2,679 29 1.09%

Mainline bet LIE EB/WB SR/43rd St entrance & Vandervoort Ave exit

4,730 4,784 54 1.14%

Exit ramp to Vandervoort Ave 920 930 10 1.09%

Mainline bet Vandervoort Ave exit & McGuinness Blvd/Meeker Ave entrance

3,810 3,854 44 1.15%

Entrance from Meeker Ave 320 300 -20 -6.25%

Entrance from McGuinness Blvd 820 915 95 11.59%

Entrance ramp from McGuinness Blvd/Meeker Ave 1,140 1,215 75 6.58%

Mainline bet McGuinness Blvd/Meeker Ave entrance & Metropolitan Ave exit

4,950 5,070 120 2.42%

Exit ramp to Metropolitan Ave 440 409 -31 -7.05%

Mainline bet Metropolitan Ave exit & Williamsburg Bridge exit

4,510 4,661 151 3.35%

Mainline bet Williamsburg Bridge exit & Metropolitan Ave entrance

2,900 2,885 -15 -0.52%

Exit to Williamsburg Bridge 1,610 1,775 165 10.25%

Entrance ramp from Metropolitan Ave 470 553 83 17.66%

Mainline bet Metropolitan Ave entrance & Williamsburg St exit

3,370 3,439 69 2.05%

Exit ramp to Williamsburg St W 890 934 44 4.94%

Mainline bet Williamsburg St exit & Flushing Ave entrance 2,480 2,505 25 1.01%

Entrance ramp to Flushing Ave 510 574 64 12.55%

Mainline after Flushing Ave entrance 2,990 3,079 89 2.98%

Long Island Expressway – Westbound

Mainline before 65th St exit 5,670 5,596 -74 -1.31%

Exit ramp to 65th St 600 443 -157 -26.17%

Mainline bet 65th St exit & LIE WE SR exit 5,070 5,153 83 1.64%



Mainline bet LIE WB SR exit & mainline/HOV split 3,120 3,086 -34 -1.09%

Mainline bet mainline/HOV split & LIE WB SR exit 2,200 2,319 119 5.41%

Mainline bet LIE WB SR exit & BQE EB/WB entrance 2,180 2,268 88 4.04%

Entrance from BQE EB/LIE WB Service Rd 760 717 -43 -5.66%

Entrance from BQE WB 360 386 26 7.22%

Entrance from BQE EB/WB 1,120 1,104 -16 -1.43%

Mainline bet BQE EB/WB entrance & Gale Ave exit 3,300 3,372 72 2.18%

Exit ramp to Gale Ave 1,260 1,142 -118 -9.37%

Mainline bet Gale Ave exit & Van Dam St entrance 2,040 2,229 189 9.26%

Entrance ramp from Van Dam St 50 42 -8 -16.00%

Mainline after Van Dam St entrance 2,090 2,271 181 8.66%

Exit to LIE WB Service Rd 1,950 2,066 116 5.95%

Entrance ramp from Maurice Ave 920 817 -103 -11.20%

Service Rd bet Maurice Ave entrance & BQE WB/48th St exit 2,870 2,883 13 0.45%

Exit to BQE WB/48th St 2,200 2,319 119 5.41%

Exit ramp to 48th St 230 222 -8 -3.48%

Exit ramp to BQE WB 1,970 2,096 126 6.40%

Exit ramp to LIE WE mainline 670 564 -106 -15.82%

Entrance from LIE WB mainline 940 818 -122 -12.98%


Service Rd after entrance from BQE WB/LIE WB 1,260 1,080 -180 -14.29%

Long Island Expressway – Eastbound

Mainline before LIE EB Service Rd (SR) exit 2,560 2,555 -5 -0.20%

Exit ramp to LIE EB SR 1,190 1,242 52 4.37%

Mainline bet LIE EB SR exit & LIE EB SR entrance 1,370 1,313 -57 -4.16%

Entrance ramp from LIE EB SR 1,620 1,752 132 8.15%

Mainline bet LIE EB SR entrance & mainline/SR merge 2,990 3,065 75 2.51%

Mainline bet LIE EB mainline/SR merge & Maurice Ave entrance

5,740 5,637 -103 -1.79%

Entrance ramp from Maurice Ave 530 614 84 15.85%

Mainline after Maurice Ave entrance 6,270 6,252 -18 -0.29%

LIE EB SR before entrance ramp from LIE EB mainline 2,520 2,531 11 0.44%

SR bet LIE EB mainline entrance & BQE WB/ Laurel Hill Blvd exit

3,710 3,773 63 1.70%

Exit to BQE WB/ Laurel Hill Blvd 430 341 -89 -20.70%



Exit to Laurel Hill Blvd 110 112 2 1.82%

Exit to BQE WB 320 228 -92 -28.75%

SR bet BQE WB/ Laurel Hill Blvd exit & LIE EB mainline exit 3,280 3,432 152 4.63%

SR bet LIE EB mainline exit & BQE EB exit 1,660 1,679 19 1.14%

SR bet BQE EB exit & BQE EB entrance 1,360 1,315 -45 -3.31%

Entrance ramp from BQE EB 2,210 2,128 -82 -3.71%

SR bet BQE EB entrance & 48th St entrance 3,570 3,443 -127 -3.56%

Entrance ramp from 48th St 310 236 -74 -23.87%

SR bet 48th St entrance & Maurice Ave exit 3,880 3,679 -201 -5.18%

Exit ramp to Maurice Ave 1,130 1,107 -23 -2.04%

SR bet Maurice Ave exit & mainline/SR merge 2,750 2,572 -178 -6.47%

To ensure that the calibration results were statistically acceptable, the volume comparisons were performed using the root-mean-square (RMS) error, the percent of RMS error, and chi-square (CS) error measures. The RMS error is a measure of the average error on individual links and hence, can be used as an overall check of the traffic assignment accuracy. Statistical theory indicates that for a particular volume group, two-thirds of the time, the difference between the assigned volume and counted volume will not be more than plus or minus the RMS error. The RMS error for each volume group was computed as follows:

N

AORMS

N

iii∑

=

−= 1

2)(

where

O = observed traffic volume on link i,

A = assigned traffic volume on link i, and

N = total number of links in a particular volume group.

While the RMS error expresses an absolute error for the volume group, the percent RMS error indicates a relative error for the volume group, derived by dividing the RMS error by the average counted volume for the volume group and multiplying by 100. The chi-square error (CS) is a cumulative measure of the error terms that emphasize large errors. The chi-square error was calculated as:

CS = ∑=

−N

i i

ii

OAO

1

2)(

As shown in Table 3, the computed error measures generally support the accuracy of the model calibration results. The value of RMS error is relatively small when compared to the mean value of traffic volumes in each volume group, implying small error in traffic assignment. For example,



for the volume group of 2,001 - 3,000 vehicles, the model has a two-thirds confidence level (about 67 percent) to assure that the difference between assigned and counted volumes is ±111 vehicles for the AM peak hour and ±89 vehicles for the PM peak hour, respectively. For the volume group of 4,001-5,000 vehicles, the model can assure that the counted volume will not differ from the assigned volume by more than ±114 vehicles for the AM peak hour and ±99 vehicles for the PM peak hour, respectively. These results are quite good for the typical area-wide assignment of this type. As also shown in Table 3, all the percent RMS errors are less than 30% for both AM and PM peak hour sub-area models. In addition, the percent RMS errors generally decrease with increasing volumes, indicating that high-volume links are more accurately estimated. The same result is also supported by the decreasing chi-square errors from low-volume links to high-volume links.

TABLE 3: MODEL VALIDATION BY GROUND COUNTS

Number of Links RMS Error Percent RMS Error Chi-Square Error Link Volume Group (vehicles per hour) AM PM AM PM AM PM AM PM

0 – 500 14 13 69.0 54.8 25.3 18.5 305.4 112.1

501 – 1,000 14 16 89.3 81.9 12.5 10.9 151.6 151.0

1,001 – 2,000 17 15 92.4 96.7 6.3 6.7 108.6 94.9

2,001 – 3,000 19 22 111.2 89.2 4.8 3.5 105.0 76.0

3,001 – 4,000 11 11 125.0 96.2 3.7 2.7 50.1 28.3

4,001 – 5,000 7 5 114.1 99.0 2.5 2.1 21.8 10.5

> 5,000 4 4 144.0 76.3 2.6 1.3 14.4 4.2

F. VISSIM MODEL CALIBRATION

Before using the VISSIM model for any analysis, a detailed calibration effort was conducted to ensure that the VISSIM could accurately replicate existing volumes, queuing, and travel times. The VISSIM model was calibrated to the base year 2002 AM and PM peak hour conditions for the roadway network shown in Figure 7.

F.1. Calibration Preparation

Prior to model calibration, the following data inputs were prepared, including:

Data Collection. Major input data required by VISSIM to develop the networks consist of four categories: roadway geometry, traffic characteristics, traffic control, and transit information. Geometric data include number of lanes, lane width, grades, curvature, etc. Traffic characteristics include vehicle classifications, speed distributions, acceleration/deceleration distribution, etc. Traffic control data consist of the locations of traffic control devices and signal timing settings. Finally, bus transit information includes routes, schedule, and bus dwell times.

Network coding. An aerial photograph of the study area was imported into the VISSIM. Scale was established on this image by matching landmarks with the scaled aerial photograph. Links and link connectors were then digitized over this background image, and various attributes from the data collection were applied.



FIGURE 7: OVERALL VIEW OF THE VISSIM NETWORK

Traffic demand. Traffic demand data consist of entry volumes and turning movements at intersections within the study area. Traffic demands were represented by two separate peak hours, AM and PM. Each peak-hour demand was partitioned into 15-minute increments to replicate the temporal variation of traffic. Within each time period VISSIM would load traffic volumes (by vehicle type) onto the entry links based on a Poisson distribution. The intersections’ turning percentages were also entered into VISSIM for the purpose of assigning vehicle movements in the network.

Traffic volumes and travel times. Traffic volume and travel time information was obtained from the project’s data collection program performed during a two-week period from November 11 to November 24, 2002. Traffic volumes were derived based on the results of ATR volume counts, while travel time data were provided by the travel time and delay runs conducted on selected routes.

Calibration criterion. The criteria used for acceptance of the calibrated models were established based on two documents: “Calibrating and Adjustment of System Planning Models” (FHWA-ED-90-015, Federal Highway Administration, December 1990), and



“Volume III – Guidelines for Applying Traffic Microsimulation Modeling Software” (Federal Highway Administration, August 2003).

F.2. Calibration Process

The VISSIM calibration is aimed at fine-tuning the model input parameters so that the resulting VISSIM model can reasonably represent the existing observed traffic conditions on the simulation network. Compared to the macroscopic traffic assignment models described above, microscopic simulation models usually contain many more input parameters to be chosen for model calibration. In the case of VISSIM, these input parameters may consist of network geometry, traffic demand, general configuration parameters, driver behavior parameters, and route choice strategies. An initial effort was given to the calibration parameters in driver behavior models (i.e., car following and lane-change models) since they directly affect vehicle interaction and govern traffic movement over the simulation network. These parameters include:

Waiting time before diffusion;

Minimum headway (front/rear);

Average standstill distance;

Additive part of desired safety distance;

Multiple part of desired safety distance; and

Ten parameters (CC0-CC9) in Wiedemann’s 1999 car following model.

The detailed descriptions of each parameter can be found in “VISSIM Version 3.7 User Manual” (Innovative Transportation Concepts, Inc. January 2003). Since the default values of these parameters were not calibrated based on the United States-based data; they were revised to reflect local traffic conditions. For example, the two parameters of “additive part of desired safety distance” and “multiple part of desired safety distance” (in Wiedemann’s 1974 car following model) were calibrated so that they could generate results as close to the maximum service flow rate of HCM 2000 as possible.

In addition to driver behavior parameters, particular attention was also paid to the vehicle parameters, such as desired speed, desired acceleration/deceleration, maximum acceleration/deceleration, and attributes associated with each vehicle type modeled. Finally, the simulation resolution was considered since it would impact the response to traffic controls such as traffic signals or priority rules.

Given all the potential parameters to be calibrated, a trial-and-error approach was adopted by iteratively adjusting these parameters to achieve an acceptable level of accuracy. In this study, the traffic counts and travel times were used as the calibration measures. Hence, the calibration target is to obtain the best match possible between the modeled traffic volumes/travel times and the observed ground counts/travel times. Based on the two documents mentioned above, the criteria for an acceptable level of calibration accuracy include:

A correlation coefficient (to be described below) would have a value higher than 0.88;



The modeled link volumes would be within 15% of the observed volumes for flows greater than 700 vph, within 100 vph for flows less than 700 vph, or within 400 vph for flows greater than 2,700 vph. These targets must be satisfied for 85% of the cases;

The GEH statistic (to be described below) would be less than 5 for individual link flows for 85% of the cases; and

The modeled travel times would be within 15% of observed travel times for 85% of the routes.

F.3. Calibration Results and Statistical Validation

The results of the AM and PM peak hour model calibration are shown in Tables 4 and 5, respectively. These tables present a comparison of simulated traffic volumes to actual ground counts for all links in the simulation network. To examine how closely model simulated data matches observed data, correlation coefficients (r2) were first calculated using pairs of simulated and counted volumes. As shown in Figures 8 and 9, both AM and PM peak hour models have a correlation coefficient of 0.99, which is substantially higher than the minimum value of 0.88 recommended by FHWA. A further comparison of individual link flows reveals that 100% and 95% of the links had simulated volumes within ±10% of the counted volumes for the AM and PM peak hours, respectively. On an “actual difference” basis, nearly all the low-volume links (i.e., less than 700 vph) and high-volume links (i.e., larger than 2,700 vph) have simulated volumes that match ground counts within ±40 and ±100 vph, respectively. These results also meet the calibration acceptance criteria described above.

TABLE 4: VISSIM MODEL CALIBRATION RESULTS – AM PEAK HOUR

Link Name

Counted

Volume Assigned

Volume Actual

Difference Percent

Difference GEH

Brooklyn-Queens Expressway – Eastbound

Mainline before McGuinness Blvd exit 4,490 4,511 21 0.48% 0.32

Exit ramp to McGuinness Blvd/Humboldt St 1,140 1,149 9 0.75% 0.25

Mainline after McGuinness Blvd exit 3,350 3,362 12 0.35% 0.20

Mainline bet McGuinness Blvd & Kingsland Ave 3,350 3,360 10 0.29% 0.17

Mainline bet Kingsland Ave & Vandervoort Ave 3,350 3,356 6 0.19% 0.11

Mainline before Meeker Ave entrance 3,350 3,331 -19 -0.56% 0.33

Entrance ramp from Meeker Ave 1,250 1,244 -6 -0.46% 0.16

Mainline after Meeker Ave entrance 4,600 4,666 66 1.43% 0.97

Mainline before Kosciuszko Bridge 4,600 4,648 48 1.04% 0.71

Mainline before exit to LIE EB/WB 4,600 4,666 66 1.43% 0.97

Exit to LIE EB/WB 2,860 2,785 -76 -2.64% 1.42

Mainline after LIE EB/WB exit 1,740 1,707 -33 -1.92% 0.80




Mainline before LIE EB/43rd St entrance 2,510 2,525 15 0.59% 0.29

Entrance ramp from LIE EB/43rd St 550 523 -27 -4.89% 1.16

Entrance ramp from LIE WB 1,930 1,902 -28 -1.47% 0.65

Entrance ramp from LIE WB/EB (merge) 2,480 2,425 -55 -2.23% 1.12

Mainline after LIE EB/43rd St entrance 4,990 4,944 -46 -0.92% 0.65

Mainline before Kosciuszko Bridge 4,990 4,935 -55 -1.10% 0.78

Mainline after Kosciuszko Bridge 4,990 4,934 -56 -1.13% 0.80

Mainline before Vandervoort Ave exit 4,990 4,932 -58 -1.16% 0.82

Exit ramp to Vandervoort Ave 960 955 -5 -0.51% 0.16

Mainline after Vandervoort Ave exit 4,030 3,974 -56 -1.40% 0.89

Mainline bet Vandervoort Ave & Kingsland Ave 4,030 3,971 -59 -1.47% 0.94

Mainline bet Kingsland Ave & McGuinness Ave 4,030 3,970 -60 -1.48% 0.94

Mainline before McGuinness Ave entrance 4,030 3,969 -61 -1.51% 0.96

Entrance ramp from McGuinness Ave 1,010 1,065 55 5.49% 1.71

Mainline after McGuinness Ave entrance 5,040 5,028 -13 -0.25% 0.18

Exit ramp to Metropolitan Ave 410 402 -8 -1.90% 0.39

Mainline after Metropolitan Ave exit 4,630 4,629 -1 -0.03% 0.02

Meeker Avenue – Eastbound

Meeker before McGuinness Ave 748 753 5 0.61% 0.17

Meeker before Kingsland Ave 952 913 -40 -4.15% 1.29

Meeker after Kingsland Ave 745 718 -27 -3.65% 1.01

Meeker before Morgan Ave 764 718 -46 -6.06% 1.70

Meeker before Vandervoort Ave 814 742 -72 -8.87% 2.59

Meeker after Vandervoort Ave 1,275 1,268 -7 -0.56% 0.20

Meeker Avenue – Westbound

Meeker before Apollo Ave 960 952 -8 -0.84% 0.26

Meeker before Morgan Ave 682 683 1 0.15% 0.04

Meeker after Morgan Ave 432 467 35 8.08% 1.65

Meeker before Kingsland Ave 494 527 33 6.64% 1.45

Meeker before McGuinness Ave 517 526 9 1.72% 0.39

Other Local Streets

McGuinness Ave SB 663 654 -9 -1.30% 0.34

Kingsland Ave NB 359 356 -3 -0.95% 0.18

Morgan Ave NB 410 414 4 1.02% 0.21



Vandervoort Ave NB 490 494 4 0.80% 0.18

Apollo Ave SB 165 166 1 0.36% 0.05

TABLE 5: VISSIM MODEL CALIBRATION RESULTS - PM PEAK HOUR

Link Name

Counted

Volume Assigned

Volume Actual

Difference Percent

Difference GEH

Brooklyn-Queens Expressway – Eastbound

Mainline before McGuinness Blvd exit 4,170 4,167 -4 -0.08% 0.05

Exit ramp to McGuiness Blvd/Humboldt St 820 844 24 2.96% 0.84

Mainline after McGuiness Blvd exit 3,350 3,319 -31 -0.92% 0.53

Mainline bet McGuiness Blvd & Kingsland Ave 3,350 3,319 -31 -0.93% 0.54

Mainline bet Kingsland Ave & Vandervoort Ave 3,350 3,317 -33 -0.99% 0.57

Mainline before Meeker Ave entrance 3,350 3,297 -53 -1.59% 0.92

Entrance ramp from Meeker Ave 1,480 1,330 -150 -10.13% 4.00

Mainline after Meeker Ave entrance 4,830 4,838 8 0.17% 0.12


Mainline before exit to LIE EB/WB 4,830 4,826 -4 -0.08% 0.06

Exit to LIE EB/WB 2,540 2,379 -161 -6.33% 3.24

Mainline after LIE EB/WB exit 2,290 2,164 -126 -5.51% 2.67


Mainline before LIE EB/43rd St entrance 2,080 2,090 10 0.47% 0.22

Entrance ramp from LIE EB/43rd St 680 682 2 0.33% 0.09

Entrance ramp from LIE WB 1,970 1,955 -15 -0.78% 0.35

Entrance ramp from LIE WB/EB (merge) 2,650 2,637 -13 -0.50% 0.26

Mainline after LIE EB/43rd St entrance 4,730 4,818 88 1.86% 1.27


Mainline after Kosciuszko Bridge 4,730 4,809 79 1.67% 1.14

Mainline before Vandervoort Ave exit 4,730 4,822 92 1.95% 1.33

Exit ramp to Vandervoort Ave 920 935 15 1.62% 0.49

Mainline after Vandervoort Ave exit 3,810 3,789 -21 -0.54% 0.33

Mainline bet Vandervoort Ave & Kingsland Ave 3,810 3,789 -21 -0.56% 0.35

Mainline bet Kingsland Ave & McGuiness Ave 3,810 3,789 -21 -0.55% 0.34

Mainline before McGuiness Ave entrance 3,810 3,790 -20 -0.52% 0.32

Entrance ramp from McGuiness Ave 1,140 1,246 106 9.32% 3.08

Mainline after McGuiness Ave entrance 4,950 5,031 81 1.64% 1.15



Exit ramp to Metropolitan Ave 440 427 -13 -2.90% 0.61

Mainline after Metropolitan Ave exit 4,510 4,609 99 2.20% 1.47

Meeker Avenue – Eastbound

Meeker before McGuiness Ave 870 894 24 2.73% 0.80

Meeker before Kingsland Ave 1,109 1,089 -20 -1.81% 0.61

Meeker after Kingsland Ave 921 863 -58 -6.28% 1.94

Meeker before Morgan Ave 921 853 -68 -7.41% 2.29

Meeker before Vandervoort Ave 977 887 -90 -9.19% 2.94

Meeker after Vandervoort Ave 1,500 1,345 -155 -10.33% 4.11

Meeker Avenue – Westbound

Meeker before Apollo Ave 920 935 15 1.67% 0.50

Meeker before Morgan Ave 746 802 56 7.56% 2.03

Meeker after Morgan Ave 587 562 -25 -4.20% 1.03

Meeker before Kingsland Ave 718 715 -4 -0.49% 0.13

Meeker before McGuiness Ave 657 622 -35 -5.33% 1.38

Other Local Streets

McGuiness Ave SB 832 762 -71 -8.47% 2.50

Kingsland Ave NB 315 339 24 7.70% 1.34

Morgan Ave NB 374 354 -20 -5.45% 1.07

Vandervoort Ave NB 438 417 -21 -4.82% 1.02

Apollo Ave SB 295 295 0 0.00% 0.00

Figure 8: 2002 AM Peak Hour Simulated Volume vs Observed Volume


0

1000

2000

3000

4000

5000

6000

0 1000 2000 3000 4000 5000 6000

Observed Volume

Sim

ulat

ed V

olum

e

`

FIGURE 8: 2002 A.M. PEAK HOUR SIMULATED VOLUME VS. OBSERVED VOLUME



Figure 9: 2002 PM Peak Hour Simulated Volume vs Observed Volume


0

1000

2000

3000

4000

5000

6000

0 1000 2000 3000 4000 5000 6000

Observed Volume

Sim

ulat

ed V

olum

e

FIGURE 9: 2002 P.M. PEAK HOUR SIMULATED VOLUME VS. OBSERVED VOLUME

To further ensure that the calibration results were statistically acceptable, the GEH statistic was computed for each individual link in the network. The GEH statistic, a modified chi-square statistic that accounts for both absolute and relative errors, is defined as:

)(5.0)( 2

ii

iii OS

OSGEH

+−

=

where

Si = simulated volume on link I; and

Oi = observed volume on link i.

As shown in Tables 4 and 5, all the GEH values are less than 5, indicating that the simulated link flows can be considered a good fit.

Travel time data was not used in the calibration process and hence can be treated as independent measurements to validate the models. To this end, travel time information was checked by comparing average travel time data obtained using the floating car method to those generated from the models within the simulation period. As shown in Table 6, travel time comparisons between field measurements and model estimation were performed for the four routes in the network and for the AM and PM peak hours, respectively. Percent differences are found to be between ±6%, indicating that the VISSIM models were calibrated very well for existing conditions.

TABLE 6: TRAVEL TIME COMPARISON

Route

Observed Travel

Time (sec)

Simulated Travel

Time (sec)

Actual Difference

(sec) Percent

Difference

AM Peak Hour

BQE EB from McGuinness Ave exit to LIE/48th St exit 163 157 -6 -3.72%

BQE WB from LIE EB/WB entrance to McGuinness Ave entrance

195 202 7 3.75%

Meeker Ave EB from McGuinness Ave to Vandervoort Ave

105 111 6 5.94%



Meeker Ave WB from Apollo Ave to McGuinness Ave 101 99 -2 -2.60%

PM Peak Hour

BQE EB from McGuinness Ave exit to LIE/48th St exit 305 297 -8 -2.64%

BQE WB from LIE EB/WB entrance to McGuinness Ave entrance

422 398 -24 -5.61%

Meeker Ave EB from McGuinness Ave to Vandervoort Ave

99 104 5 4.56%

Meeker Ave WB from Apollo Ave to McGuinness Ave 98 102 4 3.57%

G. CONCLUSIONS

The purposes of this Technical Memorandum are: (1) to outline an acceptable methodology for developing a base year model system for the Kosciuszko Bridge project, and (2) to document the model calibration process and the results obtained. The memo demonstrates that the proposed model structure was built on currently available state-of-the-art modeling techniques and hence should provide reliable forecasting information to meet the Federal mandates and standards for accuracy. In addition, extensive calibration efforts were undertaken to ensure that the base year models replicated the existing travel patterns and traffic operations in the Kosciuszko Bridge corridor. The series of checks, taken individually and collectively, demonstrate that in all respects the calibrated models (including both the BPM sub-area model and VISSIM simulation model) quite reasonably estimate the travel patterns/traffic operations on the major highways as indicated by both the ground counts and the survey data. Therefore, it is concluded that all of the techniques have been validated and are ready for use in the forecasting phase of this project.

Appendix B-1 Technical Memoranda - New York State ......NEW YORK STATE DEPARTMENT OF TRANSPORTATION KOSCIUSZKO BRIDGE PROJECT ... Flushing/Bedford Rezoning FEIS (March 16, 2001) TRANSCOM

Documents