Office of Operations Freight Management and Operations

FAF2 The Second Generation of Freight Analysis Framework

Slide 1: FAF2 The Second Generation of Freight Analysis Framework

Part 1 of 2

Office of Freight Management and Operations
Federal Highway Administration

Slide 2

Tianjia Tang, PE, Ph.D.
tianjia.tang@fhwa.dot.gov
202-366-2217
202-366-7909

Slide 3: Part 1 Outline

  1. FAF Objective
  2. Deliverables
  3. Approaches
  4. CFS Contribution TO FAF2
  5. FAF1 vs. FAF2
  6. Current Status

Slide 4: FAF2 - Objective

  1. To answer freight shipment volume and congestion questions on the highway system
  2. To answer freight shipment route/corridor questions
  3. To provide state DOTs and MPOs control freight traffic data
  4. To answer freight shipment modal questions
  5. To provide disaster scenario analyses as related to freight movements

Slide 5: FAF2 - Objective

  1. FHWA Policy and investment analyses
  2. State DOTs and local MPOs control data
  3. Other derivative usages

Slide 6: FAF is not ..

a computer model in the traditional sense of computer software modeling and programming

FAF consists of:

  1. Customized loglinear and iterative proportional fitting programs to deal with O-D matrix data
  2. Economic models to deal with future economic growth projections
  3. Customized programs to convert tonnage to # of trucks
  4. Commercial available traffic modeling programs to assign truck/passenger vehicles on to a highway network.

Slide 7: FAF2 - Products and Deliverables

  1. freight flow databases
  2. highway freight truck movement databases and flow networks
  3. waterway freight shipment databases
  4. rail freight shipment databases

Slide 8: FAF2 - Products and Deliverables (time series products)

  1. 2002 base year
  2. Annual provisional estimates starting with year 2005
  3. Projections for years between 2010 to 2035 with a 5-year increment

Slide 9: FAF2 - Approach

  1. Data Sources – public data
  2. Methods and Models – transparent and duplicable
  3. Deliverables – data, methods, and results are publicly available

Slide 10: Commodity Geographical Resolution …

  1. Observed data – “truth”
  2. Synthesized data – “trusted” or “believed”

…geographic resolution is referring to the scale of an area where “observed data” are available

Slide 11: Commodity Geographical Resolution …

  1. Observed data – “truth”
  2. Synthesized data – “trusted” or “believed”
Commodity geographical resolution is illustrated by image of a red racecar (SPEEDY) on the right at the beginning of a highway with solid and broken yellow center lines. Beginning point is labeled A and endpoint B. Interim points are labeled K and D. Observed data include that the distance from A to B is 80 miles, SPEEDY travels from A to B in exactly 1 hour, and the average speed is 80 miles per hour between A and B.

Slide 12: Commodity Geographical Resolution …

  1. Observed “truth”- 80 mi/hr between A and B
  2. Speeds between A and K, K and D, D and B can be “synthesized” based on the above “truth” and information gathered not related to the present travel.

The geographic resolution here is the length of AB and not anything shorter than AB

Commodity geographical resolution is illustrated by image of a red racecar (SPEEDY) on the right at the beginning of a highway with solid and broken yellow center lines. Beginning point is labeled A and endpoint B. Interim points are labeled K and D. Data such as speeds between A and K, K and D, and D and B can be synthesized based on observed data.

Slide 13: FAF2: Commodity Geographical Resolution…

  1. 114 Commodity Flow Survey (CFS) regions
  2. 17 additional international gateways (AIG)
  3. 7 International trade regions

A Total of 17,196,732 Matrix Cell Values for a given time period

Slide 14: 114 CFS Zones (Green and White)

Outline map of the United States shows 114 Commodity Flow Survey regions as numbered green areas and 17 additional international gateways as magenta areas. Additional international gateways include land crossings in Texas, Washington, Minnesota, and New York; ports in Texas, South Carolina, Maine, Georgia, Louisiana, and Alabama; and an airport in Alaska.

Slide 15: FAF2 - 13 AIG (Magenta)

Outline map of the United States shows 114 Commodity Flow Survey regions as numbered green areas and 17 additional international gateways as magenta areas. Additional international gateways include land crossings in Texas, Washington, Minnesota, and New York; ports in Texas, South Carolina, Maine, Georgia, Louisiana, and Alabama; and an airport in Alaska.

Land Crossings: Laredo TX; Blaine WA; International Falls MN; Champlain/Rouses Point NY; Alexandria Bay NY; El Paso TX; Brownsville/Hidalgo,TX.

Ports: Beaumont TX; Charleston SC; Portland ME;Savannah GA; Morgan City LA; Corpus Christi TX; Lake Charles LA; Baton Rouge LA; Mobile AL;

Airport: Anchorage AK

Slide 16: FAF2 - 7 International Trade Regions

  1. Canada
  2. Mexico
  3. Latin and South America
  4. Asia
  5. Europe
  6. Middle East
  7. Rest of the World

Slide 17 FAF2 – Transportation Mode

  1. Truck
  2. Rail
  3. Water
  4. Air
  5. Pipeline
  6. Intermodal
  7. Others

Slide 18: FAF2 – Commodity Classification

Categories Outside CFS

  1. Farm based agricultural
  2. Fishery
  3. Crude petroleum
  4. Natural gas
  5. Municipal solid waste

continued …

Slide 19: FAF2 – Commodity Classification

Categories Outside CFS

Continued …

  1. Construction
  2. logging
  3. Services
  4. Publishing
  5. Retail
  6. Household & Business Moves
  7. Imports

Slide 20: FAF2 – Commodity Classification

Standard Classification for Transported Goods (SCTG)

At 2-digit level, there are 42 commodities+ 1 unknown.

Slide 21: FAF2 - Regional OD Database

Example - for a given commodity and given mode with the unit of tonnage…

Image of an example table in the database. The values in the example table are not relevant.

Slide 22: FAF2 - Regional OD Database

Example - for a given commodity and given mode with the unit of tonnage…

Image of an example table in the database. The values in the example table are not relevant.

missing values and zero cell entries

Slide 23: FAF2 - Regional OD Database
The Method adopted in handling missing values

If a matrix entry is blank and it is not zero, then a value is estimated. The estimated value is then plugged back into the matrix. The method adopted here is

loglinear method

Slide 24: FAF2 - Regional OD Database
The Method adopted in handling control total

In order to maintain the control total (marginal total) of a matrix after estimated values are plugged in, a mathematical method is needed to readjust all cell values with certain relationships, the method adopted here is

iterative proportional fitting (IPF) method

Slide 25: FAF2 - Regional OD Database

Final Database

Image of an example table in the database. The values in the example table are not relevant.
  1. Unit: Tonnage and $
  2. Mode: 7
  3. Commodity: 43
  4. Geographic Areas: 138

Slide 26: FAF2 – Future Goods Movement

The regional economic growth projections for years between 2010 and 2035 are analyzed through economic models for three different scenarios – normal growth, low growth, and high growth. These growth data are commodity specific.

Slide 27: FAF2 – Routing Future Goods Movement

For a given commodity, the modal split remains constant unless information indicates otherwise.

This does not guarantee that total modal share is constant. If commodity shares are changed, then total modal share will be changed, too.

Slide 28: FAF2 – Routing Future Goods Movement

For the highway mode, routing is developed based on both truck and passenger vehicles through the least resistance assignment algorithm.

This task is carried out through commercially available traffic modeling software

Slide 31: FAF1 vs. FAF2

  1. Data Sources: private proprietary vs. public
  2. Input Data Resolution: county vs. regional/state level
  3. Output Data Resolution:
    Within USDOT: County vs. County
    Outside USDOT: State vs. Sub state
  1. State DOTs and Local MPOs Applications:
    FAF1: limited usage in modeling application
    FAF2: control total in statewide and MPO modeling
  1. Results and deliverables:
    FAF1: Limited O-D data
    FAF2: At a minimal, regional and sub-state OD data.

Slide 32: CFS Contribution to FAF2

FAF2 2002 base data, depicted as a blue oval, comes from contributors, depicted as green lines. The 2002 Commodity Flow Survey is the largest contributor, and out-of-scope studies and other data such as Waybill and Waterborne Commerce are lesser contributors.

Slide 33: CFS Contribution to FAF2

  1. CFS accounts for about 60% of total FAF2 in tonnage
  2. CFS accounts for about 65% of total FAF2 in value
  3. CFS accounts for about 71% of total FAF2 in ton-mile

Slide 35: FAF2 - State and MPO Application Illustration

Graphic showing that freight traffic in a state or local area is made up of 4 components: (1) outbound shipments; (2) inbound shipments; (3) shipments moving wholly within the state or local area; and (4) shipments neither originating nor terminating in the state or local area but passing through.

Slide 36: FAF2 – Application Illustration (daily Truck Volume)

Red lines on outline map of the United States depict daily truck volumes on interstate highways. Heavier red lines indicate higher volumes. No red lines appear on outline maps of Alaska and Hawaii.

Slide 37: FAF2 – Application Illustration (Route Freight Burden)

Inter-State Highway # No of Segs Miles FAF Trk ADT FAF Trk ADT per Mile FAF Trk Share of Total ADT
51,3041,38314,98910.822.7%
814934814,27040.969.4%
101,4652,46011,4084.629.5%
156101,4367,0144.923.8%
209561,5396,8444.420.4%
241733086,79722.019.3%
255561,0626,1085.830.4%
261682596,07123.516.0%
293737526,0618.145.4%
302863655,62915.413.3%
358971,4255,3863.816.8%
401,2182,4935,1292.118.0%
443906304,9787.918.5%

Slide 38: FAF2 – Application Illustration (Disaster Impact on Freight)

Detail Summary of Characteristics Associated with the Total Tonnage of
Freight Involving the Scenario 2 Impact Area in 2005
2005 Annual Tonnage by Shipment Mode Shipment Direction Relative to Impact Area
Into		 Shipment Direction Relative to Impact Area
Out of		 Shipment Direction Relative to Impact Area
Within		 Total Tonnage
Air 99,131 58,232 0 157,363
Highway 147,086,500 173,853,913 68,117,299 389,057,713
Other 60,819,561 31,044,424 21,322,997 113,186,982
Rail 65,154,962 34,086,432 4,304,313 103,545,707
Water 113,574,497 84,250,706 22,292,179 220,117,382
All Mode 386,734,651 323,293,707 113,036,788 826,065,146
Red lines on detailed map of coastline of Gulf of Mexico from Galveston, Texas, to Panama City, Florida, depict FAF truck flows in 2005 on interstate highways. A black circle surrounds Baton Rouge and New Orleans, Louisiana, and Gulfport and southern Mississippi. Heavier red lines indicate heavier flows. No heavy red lines appear in southeastern Louisiana or southern Mississippi.

Slide 39: FAF2 - Status

  1. 2002 base case database – January 2006
  2. 2005 annual provisional estimate – March 2006
  3. Annual provisional estimate – March of the following year
  4. Flow database – Fall 2006
  5. Projections of 2010-2035 – Fall 2006

Slide 42

End

Office of Operations