Planning Freight-Efficient Land Uses: Methodology, Strategies, and Tools (2022)

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

B-1   The proposed interaction index is based on the assumptions embedded in the gravity model [equation (1)], which have been used in transportation modeling for decades, primarily, for estimation of trip distribution and accessibility analysis. These models consider the inter­ actions between producers and attractors of trips and the travel impedance between them. The traditional model estimates the number of trips, Tij, between an origin i and a destination j, separated by a distance dij (or another measure of travel impedance such as travel time or cost). In its simplest form, population is used as a proxy of the trip production at the origin and the level of attractiveness of the destination. Its mathematical form is as follows: (Ortúzar and Willumsen 2011): = α β (1)T PP d ij i j ij Where: • Pi and Pj: populations at the origin and destination • dij: distance between i and j (or another suitable measure of travel impedance between i and j) • α, β: parameters to be estimated The version of the gravity model used here assumes that the interaction terms between zones follows equation (2). As shown, the model uses employment as the proxy of production and attraction instead of population. Thus, the interaction term is a function of the employment at zones i and j for industry sector k, and the physical separation between them. Thus, the inter­ action term between i and j is equal to: γ = (2) E E dij k i k j k ij Equation (2) implies that: (1) the larger the employment at origin i, Eki or the employment in industry at destination j in sector k, the larger γ ki j becomes; and (2) the larger the physical separa­ tion between i and j is, the lower γ ki j becomes. In numerical terms, the larger the values of γ ki j, the larger the estimated interaction between i and j (although the discussion here used total employment at i, and the employment in industry k at j, other employment variables could be used). The first step to calculate the interaction index is to compute the interactions terms between zones. These interactions terms, γ ki j, are based on the gravity model in equation (1). The second metric of interaction attempts to measure the interaction of zone i with all other zones. Since γ kij measures the interaction term between i and a single zone j for industry sector k, A P P E N D I X B Interaction Index Research

B-2 Planning Freight-Efcient Land Uses: Methodology, Strategies, and Tools the total interaction for origin i is simply the summation of γ ki j for all j (without computing the term for i = j). See equation (3): ∑ ∑Γ = γ = ≠ ≠ (3) E E di k ij k j i i k j k ijj i e interpretation of Γki is similar to that of γ kij. e larger the value of Γki , the larger the level of interaction between i and the rest of the zones, in relation to industry k. In cases where multiple industry sectors are considered in the analysis, the same logic applies. In this context, adding the values of Γki for all sectors leads to the overall value of the interaction index, as show in equation (4). If only one industry sector is being used (or if the analysis uses total employment), Γki and ΓTi are the same. ∑Γ = Γ (4)iT ik k Moreover, since ΓTi measures the interactions between zone i and all other zones, it is fair to assume that the zones with the highest values of ΓTi are likely to be part of the economic pole(s). However, determining which zones to include as part of the economic pole(s) is not straightforward. e reason is that there is no clear-cut divide between the zones that belong and those that do not belong to the economic pole. In this context, the analyses conducted by the team indicate that to ensure correct identication of the economic pole(s), it is important to add an employment constraint to ensure that the identied economic pole(s) in the study area have a sizable portion of the total employment in the area. To identify the economic pole(s), the zones were ranked in descending order of ΓTi . en, the top zones that make up at least 5% of the MSA’s total employment are dened as economic pole(s). e team selected the threshold of 5% aer experimenting with other thresholds and concluding that 5% provided the most consistent results. If the selected zones are close to each other, there would be a single economic pole. However, if they are grouped in clusters that are relatively far from each other, the impli- cation is that the study area has a polycentric structure. To illustrate these concepts, consider the case shown in Figure B-1. e example in the gure shows ve zones with dierent levels of employment. e industry sector is not considered. Figure B-1. Example of zone conguration.

Interaction Index Research B-3   The figure also shows the distances between the zones. Table B­1 shows the employment by zone, and Table B­2 contains the travel distances between all origins and destinations. Table B­3 shows the results for each. For example, for origin 1 and destination 2, the multi­ plication of 500 by 400 divided by 6 is equal to 3,333.3. As mentioned before, if the origin and destination are the same, γ ki j is assumed to be equal to zero. To compute Γki the sum of all the interaction terms must be calculated for each origin. The results are shown in the last column of Table B­3. The results show that zone 2 is the one with the highest interaction in the network, followed by zone 3. Zone 1 is a good example of a section of a metropolitan area with substan­ tial employment, but that cannot be the pole because it is simply too far away to intensively interact with the rest of local economy. To assess the robustness and conceptual validity of the proposed interaction index, the team selected six different MSAs, estimated the interaction indexes, and identified the economic pole(s). The calculations were performed using distances. The MSAs selected are New York City; Los Angeles; Houston; Washington, DC; New Orleans; and Albany. The results from the interaction index were normalized with respect to the maximum value of the interaction index for each MSA, for ease of interpretation. Figure B­2 shows the results based on total employment for Los Angeles MSA. The analyses of the geographic distribution of the interaction index results reveal four main economic poles [i.e., Los Angeles Airport (west of Inglewood), Irvine, Downtown, and Burbank]. This sup­ ports the claims in the literature that Los Angeles MSA has a polycentric form. The city of Santa Clarita was also found to be one of the economic pole(s) in the Los Angeles MSA. This result is not surprising given the fact that this city is the third largest in Los Angeles County. The next analysis considers FIS and SIS employment, separately. Conceptually speaking, treating them separately makes sense because, although all sectors of the economy interact with Distance 1 1 1 12 2 3 3 4 4 5 5 0 0 0 0 0 6 6 6 6 2.3 6.2 5.5 5.5 1.4 1.4 4.4 4.4 3.6 2.3 5.6 5.66.2 3.6 Table B-1. Employment by zone. Table B-2. Zone-to-zone distance matrix. Zones Interactionindex 1 0.0 33,333.3 25,000.0 2,173.9 2,822.6 63,329.8 2 33,333.3 0.0 120,000.0 727.3 10,000.0 164,060.6 3 25,000.0 120,000.0 0.0 681.8 2,916.7 148,598.5 4 2,173.9 727.3 681.8 0.0 62.5 3,645.5 1 2 3 4 5 Table B-3. Calculation of the interaction index. Zone Employment 1 500 2 400 3 4 5 300 10 35

B-4 Planning Freight-Efcient Land Uses: Methodology, Strategies, and Tools Figure B-2. Results for the Los Angeles MSA (total employment).

Interaction Index Research B-5   Figure B-3. Results for FIS (left) and SIS (right) in the Washington, DC, MSA. all other sectors albeit indirectly, the strongest interaction takes place among the sectors in the same group. Figure B­3 (left) illustrates the results for FIS in the Washington, DC, MSA. The results suggest that zones with the highest interaction are clustered around the downtown, which is not surprising considering the large number of Retail Trade and Accommodation and Food Services establishments in the downtown area. The area of Sterling, west of downtown, was found to have the second highest interaction index, as it includes part of the Washington Dulles International Airport; and several branches of major companies such as Neustar, Alcatel, AOL among others. Figure B­3 (right) shows the results for the SIS. In this case, the areas with the highest level of interaction are downtown, Alexandria and McLean. In Alexandria, there are several federal agencies including U.S. Department of Defense, U.S. Department of Com­ merce, National Science Foundation, among others. Companies located on the McLean area are Capital One, Booz Allen Hamilton, and Gannet Company among others. The results for the FIS in Houston MSA are shown in Figure B­4 (left). The higher inter­ action zones are clustered around the downtown area. Furthermore, the results show that the area between Jersey Village and downtown is essentially a single economic pole. The inter­ action between downtown and its surrounding areas depend on the I­10, I­610, I­69 and US 290 highways. A large density of businesses is observed along those corridors. In addi­ tion, the area of Aldine is located near the George Bush International Airport and has good accessibility to highways I­45, I­69, North Sam Houston Parkway, and Hardy Toll Road. Last, results highlight, in the east of the MSA, the connection with the ports, such as Galveston Port, and the area of La Porte where major economic hubs are located: Bayport Industrial District, Battleground Industrial District and the Barbours Cut Container Terminal. Figure B­4 (right) shows the results of the interaction index for SIS. Downtown Houston is undeniably

B-6 Planning Freight-Efficient Land Uses: Methodology, Strategies, and Tools the main economic pole for SIS. Again, the highways play a major role in determining the areas with higher interaction. The zones west of downtown between I­10 and I­69 are the areas with larger interaction. Gulfton, a zone located beside I­69, is the most densely populated community in Houston. Given the fact that a considerable number of Hispanic residents live in that area, a numerous amount of Central American and South American businesses have located in that zone, such as Grupo TACA, Famsa, Pollo Campero, among others. Conducting the analyses at the level of individual sectors, at two­digit NAICS, also provides interesting insight. Using New Orleans as the case study, the team performed the calcula­ tions for Manufacturing, and Accommodation and Food Services for the New Orleans MSA. Figure B­5 left shows the results for the interaction index for the industry sector of accom­ modation and food. The economic pole of this activity is downtown. This zone is the combina­ tion of two areas: the French Quarter and Warehouse district. The first zone is the center of tourism of New Orleans, which has several hotels, bars, stores, and restaurants. The second zone was a former logistics area focused on warehouses. These warehouses have been repur­ posed for restaurants, bars, museums, among others. The amalgamation of these two areas form the pole of this economic sector. Figure B­5 (right) illustrates the results for the industry sector of Manufacturing. The highlighted area in the east represents Chalmette Vista, where there are several industrial suppliers and refineries (e.g., Boasson Global, Rain CII Carbon and Chalmette refinery). In the south, the area of Belle Chase—which is the entrance to the Port of New Orleans—is one of the areas with a larger interaction. This zone is known for being the location of the Naval Air Station Joint Reserve Base. In addition, the area of Elmwood— at the west of downtown—is where the Strategic Petroleum Reserve of the Country is located. This facility can store 727 million barrels of oil. Given the importance of tourism to the New Orleans’ economy, it is not surprising to find that its downtown is the economic pole for Accommodation and Food and Services. In contrast, Figure B-4. Results for FIS (left) and SIS (right) in the Houston MSA.

Interaction Index Research B-7   Figure B-5. Results for Accommodation and Food Services (left) and Manufacturing (right) in the New Orleans MSA. the Manufacturing sector is scattered in the south of the MSA with multiple poles, centered around the Port of New Orleans, Chalmette Vista, and Elmwood. In some cases, due to the nature of their business, the manufacturers require a significant amount of area to produce their products and have to be located outside of the downtown area. This difference suggests that each industry sector has its own needs and preference for location. Reference 1. Ortúzar, J.D. and L.G. Willumsen, Modelling Transport. 4th ed. 2011, New York: John Wiley and Sons.