Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration
Office of Planning, Environment, & Realty (HEP)

Talking Freight

Freight Analysis Framework Forecasts and 2005 Provisional Estimates

May 16, 2007 Talking Freight Transcript

Jocelyn Bauer:
Welcome to the Talking Freight Seminar Series. My name is Jocelyn Bauer and I will moderate today's seminar. Today's topic is Freight Analysis Framework Forecasts and 2005 Provisional Estimates. Please be advised that today's seminar is being recorded.

Today we'll have three presenters. Tianjia Tang of the FHWA Office of Freight Management and Operations, Edward Fekpe of Battelle, and Charlie Han of Macrosys.

Dr. Tianjia Tang is in the position of transportation specialist with the Federal Highway Administration's Office of Freight management and Operations in Washington DC. In his position, Dr. Tang is responsible for the Freight Analysis Framework also known as FAF and the Freight Modeling Improvement Program. As the program manager, his responsibility is to ensure that the Department has a national tool in analyzing various freight related policy issues. Before joining the Freight Office in Federal Highway, Dr. Tang worked in FHWA's Resource Center where he provided technical assistance to State DOTs, MPOs and various other organizations in areas of travel demand modeling and transportation conformity. Prior to his Federal tenure, he had over ten years experience in private consulting and State DOT works. His experience covers transportation planning, NEPA, engineering design and management. Dr. Tang is a registered professional engineer in the State of Georgia.

Dr. Edward Fekpe is a transportation engineering specialist with over 25 years professional experience. Dr. Fekpe is currently a Research Leader in Battelle's Transportation Division where he provides technical leadership to the transportation policy and operations research group. Dr. Fekpe served as the principal investigator for several freight related projects for the U.S. Department of Transportation, Office of the Secretary, FHWA's Office of Transportation Policy Studies, Office of Freight Operations and Management, and Office of Highway Policy Information. He has worked on projects dealing with different aspects of transportation engineering in the United States, Canada, and abroad. Dr. Fekpe has published more than 65 technical papers in reputable journals and refereed conference proceedings and over 40 technical reports. He is a member of the American Society for Civil Engineers and serves on TRB's Vehicle Size and Weight and International Activities Committees.

Dr. Charlie Han is the president of MacroSys Research and Technology, a consulting firm based in Washington D.C. Dr. Han received his Ph.D. in Economic Geography and Regional Science from Boston University. Since 1994, Dr. Han has been working as a senior consultant to the Bureau of Transportation Statistics and Research and Innovative Technology Administration of the U.S. Department of Transportation. Dr. Han's major achievements in transportation related research and analysis includes the U.S. Transportation Satellite Accounts, U.S. Transportation Infrastructure Capital Stocks Accounts, New Measures of Transportation's Contributions to the U.S. Economy, A New Framework for Government Transportation Financial Statistics, U.S. Transportation Services Index, and Transportation Multifactor Productivity Index. Dr. Han is MacroSys project manager for the FAF provisional estimates project.

I'd now like to go over a few logistical details prior to starting the seminar. Today's seminar will last 90 minutes, with 60 minutes allocated for the speakers, and the final 30 minutes for audience Question and Answer. If during the presentations you think of a question, you can type it into the smaller text box underneath the chat area on the lower right side of your screen. Please make sure you are typing in the thin text box and not the large white area. Please also make sure you send your question to "Everyone" and indicate which presenter your question is for. Presenters will be unable to answer your questions during their presentations, but I will start off the question and answer session with the questions typed into the chat box. Once we get through all of the questions that have been typed in, the Operator will give you instructions on how to ask a question over the phone. If you think of a question after the seminar, you can send it to the presenters directly, or I encourage you to use the Freight Planning LISTSERV. The LISTSERV is an email list and is a great forum for the distribution of information and a place where you can post questions to find out what other subscribers have learned in the area of Freight Planning. If you have not already joined the LISTSERV, the web address at which you can register is provided on the slide on your screen.

Finally, I would like to remind you that this session is being recorded. A file containing the audio and the visual portion of this seminar will be posted to the Talking Freight Web site within the next week. We encourage you to direct others in your office that may have not been able to attend this seminar to access the recorded seminar.

The PowerPoint presentations used during the seminar will also be available within the next week. I will notify all attendees of the availability of the PowerPoints, the recording, and a transcript of this seminar.

We're now going to go ahead and get started. Today's topic, for those of you who just joined us, is Freight Analysis Framework Forecasts and 2005 Provisional Estimates. Our first presentation will be given by Tianjia Tang of the FHWA Office of Freight Management and Operations. As a reminder, if you have questions for Tianjia please type them into the chat box and they will be answered in the last 30 minutes of the seminar.

Tianjia Tang:
Thank you. Let us take a look at exactly what this data is about. The provisional commodity OD data is a comprehensive multimodal commodity origin destination database for the entire U.S. and includes both the imports and exports. The data set has identical dimensions to the freight database developed under the Freight Analysis Framework II. So for the moment, it has origin destination, commodity and emotional in pounds and dollars. The provision commodity of the data set is released by the end of March each year covering the most recent past year. For example, in March 2007, FHWA released the provisional data for calendar year 2006; and in March 2008, FHWA will release the provisional data for calendar year 2007.

The term of provisional data here means that the dataset is the best that can be organized at the time with the available resources; and the Freight Office has no plans to update this dataset in the future. Why do we do it? What is the reason behind the Federal Highway Administration to develop this provisional database? If you know the freight analysis framework, right now we're in the second generation phrase phase, we started the first one in 2001, the 2001, known as phase one, covered from 1998, 2010, on the 20/20. With the phase to work the base year 2002 then goes to 2010 to 2035 with the interval. From our experience with the phase one and phase two, the customers both at a federal level, state and local agencies, including organizations, they are in high demand and the more recent comprehensive national freight data. Also, we did it because of the recognition of the need by various TRB committees and it was demanded by the most recent TRB special conference on Freight Demand Modeling: Tools for Public-Sector Decision Making

Well, now, we that have the 2006 provisional commodity data online, how are you going to use it? The good thing is that the provisional data has identical format and definitions as the FAF II commodity database. So all database programs that use the FAF II are still applicable to the provisional database. All commodity GIS linkage is always applicable to the provisional database.

Who is involved? Well, this provisional commodity origin destination database is sponsored by the Office Freight Management and Operations, with a team including Battelle Memorial Institute, MacroSys Research and Technology, and University of Tennessee organized to develop the database.

Since this is a long-term effort, we're planning to do it for 2008, 2009 and years in the future. As I have said earlier, we do not have plans to update those commodity data. So we need your feedback. So, that will help us to do the next year data better. So my emphasis here is your feedback on what seems wrong in the database. That is even more appreciated than hearing about the good job that FHWA is doing. That's all I have here. I'm going to turn it to Battelle to go farther into the technical aspects of it.

J. Bauer:
All right. Well thank you, and thank you to those of you who are out there and if you have any questions please remember to post them. We'll address them at the end of the seminar and now, we'll move on to the next presentation, given by Edward Fekpe of Battelle.

Edward Fekpe:
All right thank you. My name is Edward Fekpe and I'm the PI on this project. I would like to first acknowledge the team. We have three companies: Battelle, being the lead and we're responsible for developing the provisional estimates for movements by water and pipeline and we're also responsible for putting the report together and compiling the databases for all modes. Macrosys led by Dr. Charlie Han, was responsible foe the air and highway modes. Dr. Mike Burton and David Clark of the Center for Transportation Studies, University of Tennessee were responsible for the rail mode. The goal of the project like Dr. Tang said is to develop estimates commodity origin-destination movements for 2005 and 2006 and so on. So far we have developed the estimates for 2005 and 2006. For these estimates, we used the FAF 2002 database as the benchmark.. The estimates were developed for the five modes, air, highway, rail, pipeline and water. The goal is to rely solely on the public domain databases, no private or proprietary data sources. The first task was to develop the methods or approaches for developing the provisional estimates. I'll identify the principal data sources that are used for each mode, all of which are public domain data sources. For highway, we have several of them. The surface transborder transportation database which is put out by BTS which contains commodity flows in North America and the County Business Pattern database which is put out by the U.S. Bureau of Census and many others. I will not go into too much detail on the highway mode. Dr. Charlie Han will be drilling down on the highway mode.

For rail, the main sources include weekly railroad traffic which contains information on carload and intermodal traffic for the U.S. Class I railroads, the two large Canadian railroads, a major Mexican railroad, and selected U.S. non-Class-I railroads; carload waybill sample which is a stratified sample of carload waybills for terminated shipments, ,and the surface transborder transportation database,

For air, there are two major sources of the -- for developing the provisional estimates. Office of Airline Information publishes the Form 41T-100 traffic data monthly on both a market and segment basis. The T-100 data contains information on the weight of air freight as well as origins, and destinations. The second data source which is published by the Census Bureau Foreign Trade Division provides information on international air freight movement.

For water mode, the U.S. Army Corps of Engineers, puts out information on movement of freight by water mode. This includes movements from state to state, within state, intrastate, imports and exports but it does include the value of the freight for domestic freight movements.

The primary source of data on movements by pipeline is the Petroleum Supply Annual (PSA) published by the U.S. Department of Energy, Energy Information Administration. This data source contains information on crude oil and petroleum products moving among the Petroleum Administration for Defense or PAD districts. The country is divided into 5 PAD districts and the data just shows movements of these products among the five districts. It doesn't show the movements between the states in each district. That's one challenge for us - to expand the PAD district movements to state level.

Having identified the principal sources of data for each mode, let me briefly discuss some of challenges that we faced. The data are not consistent as one would imagine from different data sources. Data were collected by different agencies using different methods and presented in different formats. We had to deal with inconsistencies in data. In some cases, some pieces of data were not available, for example, for the domestic air movement put out by the Office of Airline Information on Form 41T-100 excludes information for some all-cargo carriers. For rail, for the public use waybill data, the origin and definition were removed. Also, for pipeline mode , the data is available by PAD districts so there's no information on how much crude oil or petroleum products are moved by pipeline between states.

One of the problems we faced was the cross walk between commodity codes. For example, the data on water movement in the public domain, the code that was used to describe commodities was different from the standard classification of transported groups. The SCTG codes were used in the 2002 FAF database, so we had to develop a cross walk between the commodity codes. Furthermore, the data was not available in the FAF region level, as such we had to expand the state to state level data to the 131 FAF zones. This consists of 114 CFS O-D zones and 17 major ports, border crossings, and freight ports so we had to expand the data, from state level to the FAF zone level to be able to be consistent with what is in FAF2 database. We also had to [calibrate the estimation models to ensure that they generate reasonable estimates. Those were the major challenges we encountered in developing the provisional estimates.

The next slide shows the methodologies. What we tried to do is to ensure that the methods that were used to develop the estimates were specific to each mode and that the method that we chose was determined primarily by the nature of data that we have. Also different approaches were used for domestic versus international movements. Just to give you an idea of the methods, we use growth rates for both state level and FAF region level estimates. I believe Dr. Charlie will give you more examples of those. For some modes, we used simple moving average approach because that approach appeared to give more reasonable estimates when we're doing the calibration. In cases where we do not have information on the value of the commodities for 2005 or 2006, we have to rely on weight-to-value ratio developed from 2002 FAF2 benchmark database and using the CPI to account for inflation.

The next slide shows the architecture of the estimation process. The first step was to identify the data sources and access the data. Then we clean up the data and try to address the challenges that I mentioned earlier. We made sure that the data is usable and has all the variables and parameters that we need and once we have the usable data, we develop the methodology based on the data that we have. The methodology was then used to generate a preliminary set of estimates. These would then be compared with the 2002 FAF2 benchmark data. We want to be sure that the estimates are consistent in the sense that estimates for 2006 are consistent with 2002 bench mark database. For example if the tonnage for a given mode and O-D pair in the 2002 benchmark database is 100,000-tons and our estimate is, say, 5,000 in 2006, then there's a question. We have to go back and look at it and make sure we get the values that are consistent or reasonable enough that it didn't jump out at anybody that is looking at the data. So we went through a series of quality control exercises to make sure that our estimates are consistent are look reasonable compared to the benchmark values. We went this quality control process for each mode. Then we generated the provisional OD databases, there are four major databases. The first one is the domestic database, that is, all movements originating and terminating within the United States. The second one is the international movements via land border crossings,. The third is international movements via seaports - that includes movements imports and exports between the U.S. and other countries. The fourth database is international movements by air. The values in brackets are the number of ) records in the 2006 database.

The next slide shows the structure of the origin-destination database. For each record, we have the origin state, origin FAF zone, and destination state and destination FAF zone. This is followed by the description of the commodity, the mode, the tonnage in kilo tons, and the value in million dollars. Some states have only one FAFA zone, like Montana, while other states have more than FAF zone. The state of Alabama has more than one FAF zone and one of them is "remainder of Alabama".

The next thing we did was to develop state and national summaries. I'll show you an example of the structure of the state summaries. Each state summary has four sheets or-- four tables for each year for which estimates were developed. The first table shows the total weight of commodities moving by each mode, originating from the state, destined for that state, and within the state. The second table shows and the total value of the commodities by mode just as the weight. The third table shows the top 5 commodities originating from the state, destined for that state, and within the state. The fourth table shows the top five trading partners with the state. The table on the screen, as an example shows tonnage of commodities moving by truck and percentage of total within the state, from the state, and to the state. We have a similar table for value as well.

What are the lessons learned? We are quite familiar with the nature and nuances of the various data sources used in generating the estimates. The methodologies we used were calibrated and tested for 2005 and 2006 estimates. We are now beginning to work on 2007 estimates. We have developed SQL queries to help conduct quality checks on data. We cannot guarantee the quality of the provisional estimates. We did best that we could do. There are always problems and issues that arise in compiling large databases -and this one is no exception. Given the limitations of the quality of the data from different sources and taken into consideration the time and the budget of available for the project, we did the best that we could and we're sure the provisional estimates are consistent with what you have in the benchmark, 2002 database. Like Dr. Tianjia Tang said, provisional estimates will be generated every year. It is important that these provisional estimates are not competing with private industry. The advantage is to give you the bigger picture and one can go to private data sources, if you want more details. We expect that the quality of the data to improve with time as we gain more experience in developing these estimates. If you have any comments or suggestions that will help us to achieve that, we would appreciate that. You can send your comments to us and we will be happy to consider them.

J. Bauer:
Well, thank you, Edward. We'll now have our final presentation given by Charlie Han of Macrosys.

Charlie Han:
Tianjia and Edward talked about objectives of the project and also covered in details all that is going on and also the meaning and the data we use to develop the provisional estimates for the highway movements. So, for the highway movements, we use approach, we characterize updating approach. This approach produces provisional estimates as the sum of the highway 48 moment highway freight moment and the changes of growth from the bench mark year to the estimate year. So I don't know whether you can see the movement of my cursor, but can you see if the bar on the bottom of the slides, between Chicago region and the New York region, if we between the bigger add this together provisional estimate of the new 48 tonnage for the estimation -- estimate per year. Okay? Freight movement information 2002, bench mark data so our primary task in this project is really to estimate the growth inspect the -- growing instead for each of -- OD region between bench mark year and estimation year. Okay. Has the following three major advantages. One is that it [Speaker/Audio unclear] full of it realizes all relevant information including most recent data to allow the most recent estimates to capture the bench mark year. This change including the structure change, the commodity component change and the reeling growth the second advantage is that its bag retains the detailed information embodied in the estimates of the bench mark year. So as we know, when FHWA developed the 2002 bench mark database, there's a lot of detailed data that ask those data are not available for the provision estimates, for example, the CFS, commodity fluctuate for 2002, you know, is the primary data input for the bench mark -- staff bench mark, database development but CFS is not available for 2005 and either for 2006. So you understanding our estimates we are -- we have to utilize that the information embodied in the 2002 bench mark year data. And also, database, there's a lot of extra work put into that effort. A lot of knowledge, a lot of experts from if HWA and from the priority private industry all work on that. So there's a lot of knowledge or extra knowledge that are embodied in that you know, database. So we want -- you know, our estimates, we want to take advantage of that additional information and also the knowledge as you're base. The third is that estimation process to the growing or change of the flow. Or the freight movement rather understand the freight movement itself. Okay so this -- you know, because we're estimating a portion of the total provisional estimate. So we hope the smaller. What we really upon is from this project is freight flow commodity and fast of the regions. This slide shows you familiar we have a fast , between any pair of this regions. Okay? Basically the flow by commodity. So, in the following, I'm going to discuss our approach. You know, previous slide, we -- we -- you know, we talked about what we really upon but we don't have that information available to us readily. Now, we're going to estimate it -- those automatic you know, information we need. The first step, you know, our approach, is to estimate the growth, national highway -- national domestic highway freight tonnage, so we're calculated this growth as the total domestic highway freight tonnage of bench mark a year times the growing rate of highway freight calculated from American trucking associations trucking tonnage index. Okay? The calculation is pretty simple. The only question you know, you may ask is Why use American -- trucking association as trucking tonnage index data? Because that's one direct observation and based on your past working experience, we see all the data and the freight data we realize that -- this is most reliable data source public available. -- publicly available. So the suspect pretty simple. Once we calculated the growing rate applied to basic years highway freight tonnage, then we got the national although -- national to the am growing -- to that growing, however, we know the total growth does not have the demission, also does not have the commodity details. So our next step is bring the total growing into growing by commodity. The method is -- commodity. The method is pretty simple, again, the growth by commodity will equal to the total growing time the share of pod it'd in total output. The total output here is measure indeed dollars. Okay? We do these because of to the tonnage in terms of tonnage, we don't have commodity details, but in terms of output in dollar values, we do have information from the census bureaus monthly survey of manufactures shipments, sending orders. -- and orders. Okay? I'm having some difficulty in moving some slides. So once we have the commodity by the highway free movement, by commodity, then our next task to distribute these growth in highway freight by commodity to in the national level to the staff region, for example. In the slides, you can see the delta X between the represent the flow commodity flow -- commodity flow, X from between region B and C. Okay? Then this is to the regional level now. So it's a course growth in highway freight off of commodity X. Between this -- not national level times the distribution factor for commodity it'd X and for region B and a C. So concept you willingly, the method is again, very trait forward, however, the challenge is really in the estimation, how we come up handily, a bad distribution factor.

So we have to come up with good distribution factor with different fact regions. We have a two steps -- two-step approach to calculate the distribution factor in two regions, we use the Delta Y as the symbol for these on -- potential growing. In my slides, you guess delta Y underscore B C rep potential growing for movement between region B and region C. So as you can tell, the -- we use TDP represent regions economic size so the delta TDP really reps the Greg between the -- represents the Greg between the bench mark year and the estimation year. So we calculate the growth between the combined growth and region C. Divided by of region B and C a year. Then we times the bench mark a years to flow X between the B and C all this is available as you can tell. The only new information we need is delta TB and C and B.

Okay. The distribution factor, the DF under score X and a BC total potential growing coming that is the sum overall the regions, the free to flow, free to movement growing. That's give you the sheer that's what we call its distribution factor. So as we know, this distribution factor for all the OD -- regions, will sum up one. Once we have this distribution factor, the calculation for provisional estimates for highway freight movement, it's pretty simple. You know, we showed in a previous slides. The is suggest the total growth multiplied by the distribution factor each year, staff of regions. Next slides.

Okay. When we have the regional growth, then we have the provisional tonnage estimates as to the bench mark years, freight tonnage, plus what we just calculated, the growth bench mark year and estimation year for each of the staff OD regions. So that will give us the provisional tonnage estimate. That's complete our estimation for the domestic highway how ferment. Okay next slides. different available to us. So here, you know, from the primary data source, as you know, talk -- Edward talked about, from the international ETS, transport database, so each staff region by commodity, that's what we have, we also have the exports from each staff region by commodity. And we -- we also have imports in terms of tonnage to -- and value through each port, the total imports through each port. Okay. And also, export through each port and but -- but, we don't have them -- this information, the imports through each ports we also know, in addition to the each ports, you know. The relationship between information you have. , that is the total auto-imports to each if you sum them up, they should equal to the imports that went within through each of the ports. Okay sum of the order exports through the ports. So, however this does not give information we want for international flows that's your OD pairs staff regions that's so we don't have that information. Okay. However, this information is available for you know what we need, have available for the staff bench mark year because they have detailed data. the information we have and that was it. The relationship .

If the following description, I use export as an example, the symbol reps export, the subscript X reps era to P one and then the T represents the year, basically. So that symbol tells us the export of commodity X from region A to part one X from region A top the estimation year without part information that's what we have. Times the share of part one inspect the exports of X from region one in a bench in year. -- bench mark year. This information can be derived from the bench mark data. The following I basically present how we calculated the sheer from part one and export from commodity X from staff region A in a bench mark year. And so it equals to the export of commodity X from region A to portion in a bench mark year divided by the X -- export of commodity it'd to that -- commodity total from region A in a bench mark year. So -- and also, the mathematical quality of this shears, for a region, if you sum up across all the ports, okay, they also should equal to one, that's the concept. So -- concept. So for commodity, not in a bench mark year because we run into the situation where you know, they call developed and there's a new product coming into the market and then that you believe so for those commodities, we don't have this detailed information in the -- even in the bench mark year, for those, we calculate with distributed them to the ports to using the average -- the average export, basically to distribute the commodity export from a region to the different ports. Okay? Once we have these export from a staff region to different ports, then we have to check to see if the sum of these export flows flows distributed at two the -- distributed to the different ports weather meted the knowing information, that's from the transported database, for example, we know the export went through each ports, right? So now, we already -- we also have the distribution of all exports by commodity and region to different ports. So if we some older distributions to a port, exports it to a port, then it's a conceptually -- it's assured the total of the -- total export that's recorded in the transported database. But as we know, we are doing the estimation, most likely this two numbers to will not believe the same. So then we calculate the differences between our estimated and what's observed inspect the transported database for each port basically, the export total export tonnage within through that port. Okay, we calculate these outside the -- using the at the bottom of slides. Basically, we sum up all the regions exports to the same port. That equals to the ports total for each tonnage for export. Next slide.

Then we need to distribute -- because the differences will be so the sum of that differences will be zero, this is quality come from that exports from all regions added together should add up to the exports through -- exports through all the ports. So then for -- for the ports, we suppress and chatted modification factor to take away the suppress from it. To make the refund estimates, sum up two total tonnage for exports for a port reported in a transport database. So we use -- we calculated these modification factor as to the actual export tonnage through a port, divided by the sum of our estimation distributed to that port to all regions. Because as we'd, the difference is -- we'd, the difference -- as we said, the difference is smaller the modification somebody smaller than one. With, this then we Multiply the mod I have gone factor we saw with initial estimate. We got refund estimate and that's estimate if we sum over the commodity it'd and all the fact regions, then we're quality to the four part distribute to the same part where you actual although -- export tonnage within through that -- that part. Okay? So the first [Speaker/Audio unclear] is for the export to port that has with our initial estimate. For the ports that has a deficit, we use a different method. We first calculate the sheer of a port that's with a deficit it's a sheer yelling the total deficit with all the portless with a deficit. Then, we are calculate the refund estimated export from you know, a region to a port as the initial estimate for that port plus the modification. The modification, as you can see from the second equation from the bottom is the sum of all the -- bottom is the sum of all the from the ports that have a sum them together that distribute according to that here to the deficit. So that give you the last equation. Then, when this estimate, because it's modified week made sure the sum of the export X from all regions to port we are quality to export of that commodity went through that port, a certain port at concern. And our provisional estimate pet one -- basically, the condition that maintained the consistency in the data we have. Okay, when that, I conclude my presentation. I hope we have some time so I can answer some questions. Thanks.

J. Bauer:
All right well thank you Charlie. I hope everyone enjoyed these presentations. And now we'll start the Q and A session. Okay our first question was what is the website to access the freight provisional database and I answered that lower down in the chat so you find that address there.

Now, here's the question from Julian. How many FAF zones are there in the nation? How many in Illinois and New York?

I believe I said it. We have 131 FAF zones three in Illinois and five in New York.

J. Bauer
Okay. Great. All right and let me go to the next one here, this one is for Tianjia. I think that you mentioned that FHWA was NOT going to update something in the future. I am not sure what is not to be updated. Can you elaborate?

T. Tang:
I'll give you one example here, it was U.S. Army Corps of Engineers, you know, they publish among the cargo data. At the end of year, typically six or mated months, prior to the end of calendar year, they will redo, recompute and reclassify, do an update on those among the charts and do a new national summary statistic says. What we are doing here is once we published though provisional data by the end of March, we no plan --we have no plans to reprocess those data, Army Corps of engineers any other data sources, are or updated them so that is why I said we have no plan to update a provisional number, that's way mean.

J. Bauer:
All right. Thank you. The next question is for Dr. Han. How are the movements of containers from ocean ports via rail to inland destinations classified? As domestic or international moves?

C. Han:
I'm not on responsible for the highway movement, the way not have the information for -- we did the work.

Let me go ahead and Charlie? You know, clearly, this is an international shipment so it is in our international sea file. But it will be classified as interpreted under model classification.

J. Bauer:
All right. Here's something from Chip Millard. What methodology was used to determine which metropolitan regions were treated as individual FAF regions? Some moderate-sized metro areas are major freight hubs. For example, Harrisburg, PA is one of Norfolk Southern's 3 primary intermodal hubs and the region is also a major truck-based warehouse/distribution center hub.

I think I'm going to take on this one. Chip, this is a really good question. You know, and have the time, I would like to redefine all those staff regions every year. But to maintain a we are using the defined in which is further define by the 2002 commodity flow . Basically, one of the 14 includes 64 major met profit than, [Speaker/Audio unclear] the 50 remains of the states or of the of the state that and that's how we came up with one of 14 regions.

J. Bauer:
All right and here's another question, this one's for Charlie Han. Does the model estimate intrazonal freight movements?

C. Han:
Is there -- we didn't do that, but we didn't because we do have the doubts economic data. It's the -- level. However, the -- we don't have the information for the bench mark year because it also only have the freight movements between the staff regional level, not within the staff region, am I correct?

Charlie, let me rephrase the question. I in your answer there is not what we have done. You know, we have conversation of everything here. We have intraregional flow, such as Alabama, the remainder of Alabama to .

Region of Alabama, so we do have intraregional flow. I saw on the screen, there is another related question to this one if you are asking us do we identify, do we motto the road for a moment if Carle's motto in this provisional, origin, definition, database exert we do not identify but we do in this database.

J. Bauer:
All right and here's a question from Caroline Marshall for Tianjia. It seems that the analysis focused primarily on modes and between regions. How was multimodal movement of freight addressed? Also, was there any analysis of intra-regional flows for any of the FAF regions?

T. Tang:
That is another good observation here. Clearly, our presentation today is geared towards intraregional movement so you do in the see this here we also did intraregional flow here. Because we do a database we might have intermodal. You know, motto, we are we have want to rail. We have motto. The only thing here which is not classified under modal is under the sea file because all those goods came from a foreign country that through water and once get out port, they're hauled away by either want to rail, want to or water again. But we're kind of dis the international water lag so how are things and one, get out of the port since that Saturday mode we are using there.

J. Bauer:
Okay. And now here's something for Dr. Han. Can the model estimate freight flows that travel through a metropolitan region (i.e., do not begin or end in the region)? If not, are there any plans to model such flows?

C. Han:
Okay, basically this is -- go through a bridge, like the bridge region, if the metropolitan region because enemy NFF continues, there are some information on the net work basically that will give you information through a state through a staff region and to another staff region.

tell cannot be answered by this provisional facial but it can be done then that that.

J. Bauer:
And are there any plans to model such flows?

C. Han:
Right now, we don't have any plans to do that model.

J. Bauer:
Here's a second question from Caroline Marshall. How are movements between multiple regions considered? I'll open that up to anyone who would like to address that am.

I'll give it a shot first. Caroline, you know when we talk about good movement model -- multiple regions but it didn't matter how many is being involved, as long as it has a two origin and two definition it may pass quite a few points what this database captures is a two-origin and a two definition. -- destination. So from the production end and the con assumption end.

J. Bauer:
It looks like we've gotten through all the questions that were posted on the online chatted. Why don't we have the operator now open up the phone lines for questions.

Thank you. If you would like to ask a question, please press star then one. You will be prompted to record your name. To withdraw your request, you may press star then two. Once again torque ask a question, please press star one and record your name. One moment.

I was just wondering if any of this information has been transplanted into numbers of vehicles. vehicles. Thank you.

The question is have those numbers in the database converted into the number of the trucks. The answer suspect no. They are still measured in pounds and it has not been converted into trucks for the highway mode yet.

J. Bauer:
All right. Thank you. do we have any others in line?

At it time, I show no audio questions.

J. Bauer:
All right. Well I think we'll wrap up a few minute says early then. We appreciate all your participation and thank you for attending the seminar. The recorded version of this event will be available within the next week on the Talking Freight website.

The next seminar will be held on June 20 and is titled "The Effects of Construction Zones on Freight Movement." If you haven't done so already, I encourage you to visit the Talking Freight Web Site and sign up for this seminar. The address is up on the slide on your screen. I also encourage you to join the Freight Planning LISTSERV if you have not already done so. Enjoy the rest of your day!

Updated: 3/29/2011
HEP Home Planning Environment Real Estate
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000