National Academies Press: OpenBook
« Previous: Chapter 2 - GA Forecast
Page 12
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 12
Page 13
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 13
Page 14
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 14
Page 15
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 15
Page 16
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 16
Page 17
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 17
Page 18
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 18
Page 19
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 19
Page 20
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 20
Page 21
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 21
Page 22
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 22
Page 23
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 23
Page 24
Suggested Citation:"Chapter 3 - Air Taxi Forecast." National Academies of Sciences, Engineering, and Medicine. 2009. Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast. Washington, DC: The National Academies Press. doi: 10.17226/14301.
×
Page 24

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.

12 Introduction To develop a fleet forecast for air taxi services involving next-generation small aircraft, the primary focus relies on the view that the demand for aircraft will ultimately be derived from the consumer demand for such services. This approach is quite different from the GA analysis of the previous chapter where the forecast depended importantly on supply factors re- lated to the financial outlook and production capabilities of the major VLJ manufacturers. The analysis that follows is best thought of as a market “potential” forecast and, because the air taxi market is just now emerging, it is inherently somewhat speculative. Because the focus is on potential market demand, it is implicitly assumed that manufacturers will be able to expand production as needed to meet the demand. The consumer demand for air taxi services can be viewed as a part of overall travel demand by consumers. The primary components of travel demand that are relevant for the current analysis include trip generation, mode choice, and trip distribu- tion. Trip generation refers to the overall number of travel trips and reflects the initial decision about whether to travel. Mode choice refers to which mode will be used for the trips. Trip dis- tribution refers to where the trips will occur; for present pur- poses, the primary interest is only in the origin and destination of each trip as opposed to the actual routing. Standard analysis of travel demand recognizes that it is a derived demand—people travel not because they enjoy travel- ling, but because it is a necessary component of some other end-use desire (e.g., meeting with clients or other business associates, going on a vacation, visiting relatives, etc.). Because of this, one cannot conclude that a newly available mode of travel (e.g., VLJs) will necessarily increase the overall demand for travel. Rather, it is more likely that VLJs may “steal” traf- fic from other existing modes of travel. Overall growth in travel demand and trip generation is likely to depend mostly on demographic trends in population and income. For the present study, a constant per-capita trip rate is assumed for the domestic United States, and overall growth in travel trips is assumed to be proportional to population growth. As discussed in more detail below, income effects are treated as determinants of travel mode choice rather than overall trip generation. Normally, one might expect that air taxi services offered on next-generation small aircraft would compete primarily with existing air taxi services—those offered by small piston, turbo- prop, and/or light jet aircraft. However, many analysts believe that air taxi services provided by next-generation small aircraft also have the potential to compete for trips that are currently taken via commercial air service and/or automobile. These lat- ter categories provide many times more trips than current air taxi services, so it is important to include these travel modes in the analysis.10 However, it also is important to recognize that not all commercial air and automobile trips will be relevant. In fact, it is likely that only a relatively small portion of such trips— in particular, business trips by high income travelers over rela- tively short distances—will be realistic candidates for switching to VLJ services. The specific filters used to restrict the “universe” of potential VLJ travelers on each mode are discussed below. The primary analysis approach used is a mode choice analy- sis. The basic approach involves defining the “universe” of potential existing trips for each mode in which next-generation air taxi services may be able to compete. Each mode is defined by a set of attributes relevant for travelers deciding among the available choices. Typically the primary attributes thought to affect mode choice in transportation studies are price (i.e., cost) and travel time. In addition, characteristics of the individuals making the mode choices (typically income) may be important. For the present analysis, it was also important to consider the impact of party size on the mode choice decision. For example, C H A P T E R 3 Air Taxi Forecast 10White papers appearing on air taxi startup DayJet’s website specifically mention the potential demand for their services from business travelers who currently travel via automobile because of the relative lack of commercial air service between many smaller destinations in DayJet’s primary service area.

suppose a group of four travelers has chosen to travel on a six- passenger light jet; statistical estimates of the likelihood of choosing a three-passenger VLJ if it were available may then depend critically on whether the group of four is travelling together and whether the light jet or VLJ services are sold on a per-seat basis or a traditional charter basis. This is discussed in more detail below. An important requirement in estimating a mode choice model is that the attributes of all modes that are available (and not just the one that was actually chosen) must be measured. The essential output from the mode choice model gives statis- tical coefficients for the mode attributes and individual char- acteristics that can then be used to estimate the probability that the individual will choose each available mode. These prob- abilities then can be translated into “shares.” For example, suppose there are 1,000 observed trips involving commer- cial air as the mode of travel; this means that 1,000 individ- uals actually chose commercial air as their preferred mode. The statistical model will generate predictions about the probability that these trips are taken by each of the available modes. It may indicate, say, an 80% probability that these trips will be taken by commercial air; an 8% probability for the automobile mode; and a 4% probability for each of the three currently available air taxi modes (piston, prop, light jet). Multiplying the probabilities (shares) by the number of trips yields projected trip totals for each mode (i.e., 800 com- mercial air trips, etc.). Then, to simulate the impact of the entry of VLJs into the market for the forecast years 2012 and 2017, a new “mode” is added with particular attributes representing VLJs, and the shares are recalculated based on the estimated coefficients. To account for generic growth in travel over time, the overall num- ber of trips is grown for the forecast years 2012 and 2017 based on population growth projections assuming the overall per- capita trip rate remains constant. Finally, estimates of passen- gers per flight and annual aircraft utilization rates are applied to transform these projected VLJ trips into fleet forecasts. For present purposes, it was much more efficient to use ex- isting survey data rather than to design and undertake a sur- vey from scratch. For the existing air taxi and commercial air modes, the best available data are those from mode-specific datasets: daily Enhanced Traffic Management System (ETMS) traffic in FY2007 collected by the FAA in the case of air taxi and quarterly Origin-Destination Survey (DB1B) for FY2007 collected by the U.S. DOT in the case of commercial air. For automobile traffic, the potential universe of trips is drawn from the 1995 American Travel Survey (ATS) conducted by the U.S.DOT. Although this survey covers all modes of per- sonal transportation (including commercial air and charter travel), it does not provide nearly the same level of geographic detail that the ETMS and DB1B datasets provide. However, it does provide a much larger sample of long-distance trips and more useful information on trip origins and destinations than its successor survey that was conducted in 2001.11 Data Details Airport Data The universe of potential airports for handling VLJ activ- ity was restricted to public-use facilities in the lower 48 states with at least one 3,000-ft lighted runway and jet fuel availabil- ity. For air taxi use, FAA medium- and large-hub commer- cial service airports were excluded from the database based on observed usage patterns from various air taxi operators showing that such airports are avoided (presumably to avoid airside and/or landside congestion at these facilities). These restrictions resulted in a “VLJ airport” universe totaling 1,842 facilities. This list in fact includes a combination of commer- cial service, reliever, and GA airports; it is meant to represent the airports that are most likely to be impacted by growth in the activity of VLJs and similar aircraft. It is likely that owners and operators of next-generation equipment will want to take advantage of the advanced avion- ics packages in their aircraft; this suggests that airports with precision approach capabilities will be most attractive to these users. In addition, airports with other amenities such as hangar facilities, ground transportation services, de-icing and snow removal capabilities, mobile auxiliary power units, and so forth will be attractive to VLJ air taxi operators. While sufficient data on these latter attributes for the 1,842 identified facilities are not available, it is possible to assess airport “readiness” for VLJs based on observed characteristics and some proxy measures. Table 7 breaks out the airports regionally based on the avail- ability of at least one precision runway, plus the number of based GA jet aircraft. It is reasonable to presume that airports that have precision approach runways and higher numbers of based GA jet aircraft are more likely to be “VLJ-ready” than those that do not. As seen in the table, the highest number of airports with precision approaches and higher numbers of based jet aircraft are in the Southern, Southwestern, Eastern, and Great Lakes regions. It is not surprising that many industry observers expect these areas to attract the highest number of VLJ oper- ations, and compatible assumptions are made below in the statistical analysis where projections of future operations are estimated. For each VLJ airport, the two closest commercial airports (those with at least three daily scheduled departures) were 13 11Clearly there is a need for more recent survey data—not only for this study, but also for work in other areas. In 2008, U.S.DOT launched the most recent National Household Travel Survey; data will continue to be collected through the Spring of 2009, and the first set of results is expected to be available late in 2009.

identified in order to facilitate a search for attributes related to the commercial air mode for each trip. A total of 416 dif- ferent commercial facilities were identified, distributed as shown in Table 8. (Note that in some cases, a VLJ facility could simultaneously be identified as a commercial facility.) Census Data One of the most important practical issues to address is the physical location of travel trips. For the current analysis, it is necessary to assign trips projected to be taken by the VLJ mode to the specific airports that are able to accommodate such flights. Demographic data at the census tract level were obtained from Applied Geographic Solutions (AGS). This dataset in- cludes estimates of population and income for 2007 and 2012, as well as population projections for 2017. Catchment area de- mographics for each VLJ and commercial airport were then identified by finding the nearest such airport to each census tract. The data was combined with the airport data to obtain catchment area estimates of current and projection-year pop- ulations and incomes applicable to each VLJ and commercial airport. Aggregations of population and income were also made to the Metropolitan Statistical Area (MSA) and at State levels to allow for projections from the automobile traffic data in the ATS. MSAs are collections of counties, cities, and other smaller defined geographic areas that together compose a sin- gle metropolitan area. Current Air Taxi Population The population of current air taxi activity that potentially could be affected by VLJ competition was derived from the ETMS data collected by the FAA for the period October 2006 through September 2007 (FY2007). The ETMS system collects individual data on all flights that enter the domes- tic en route system. GRA, Inc. has performed an extended analysis of air traffic for the FAA using this data; part of this analysis categorizes each flight into a user category based on N-number, aircraft type, and owner or operator identifica- tions. GRA identified six different user groups composing passenger charter and/or non-scheduled Part 135 passenger operations. The flights in these categories were then further trimmed by applying the following filters: • Departure or arrival airport must be in the list of “VLJ- ready” airports identified above, • Size of the aircraft must be between 3 and 8 seats, and • Great-circle flight distance must be between 150 and 800 mi. The logic behind these filters is that the introduction of a new VLJ alternative is likely to attract only those existing air taxi passengers who already are flying to VLJ airports, on air- craft of similar size to VLJs, and within the non-stop flying range of VLJ aircraft. A total of 146,763 annual flights were identified with this methodology, and each was assigned to the piston, turboprop, or light jet air taxi mode. It is important to recognize that the ETMS data essentially covers only those flights operating under Instrument Flight Rules (IFR). To account for the potentially large amount of cur- rent activity that operates under Visual Flight Rules (VFR)— which is not in the ETMS data but that nonetheless might be captured by VLJ air taxi services—the flight counts were then scaled up using data from the 2006 General Aviation and Part 135 Activity Survey conducted by the FAA. Specifically, activity ratios of total air taxi flights to IFR air taxi flights by engine type (jet, prop, or piston) were computed and applied to the filtered ETMS data. This resulted in estimates of current relevant air taxi flights as shown in Table 9. 14 No Based GA Jet Aircraft 1-10 Based GA Jet Aircraft 11+ Based GA Jet Aircraft No Based GA Jet Aircraft 1-10 Based GA Jet Aircraft 11+ Based GA Jet Aircraft 16625111132285Central 16750231034050Eastern Great Lakes 172 1 74 34 33 73 387 New England 15 1 10 15 6 10 57 Northwestern Mountain 108 5 40 10 13 36 212 4036037181198161Southern Southwestern 147 3 65 13 28 43 299 Western Pacific 59 16 38 6 13 19 151 184231616411741236797Total Number of Airports WITHOUT a Precision Runway and: Region Number of Airports WITH a Precision Runway and: TOTAL Table 7. Distribution and attributes of VLJ airports. Region Count 31Central 57Eastern 72Great Lakes 22New England Northwestern Mountain 68Southern 54Southwestern 47Western Pacific 416Total 65 Table 8. Distribution of associated commercial airports.

The ETMS observations are flights that are shown on an airport-to-airport basis; in order to calculate total trip time (which should include airport access and egress times for the commercial air and air taxi modes), it was necessary to distrib- ute the passengers on these flights to surrounding areas. As noted earlier, this was done using Census population estimates to generate synthetic observations of air taxi travel at the Cen- sus tract level. The conversion of ETMS flights into passenger trips is discussed below in the Model Assumptions section. Current Commercial Air Population The population of current commercial air traffic that poten- tially could be affected by VLJ competition was derived from the Ten Percent Origin-Destination Ticket Sample (DB1B) col- lected by the U.S.DOT for the period October 2006 through September 2007 (FY2007). Unlike the ETMS dataset, the DB1B data is measured in passenger trips directly, not flights. As an initial screen, only domestic trips within the lower 48 states and between 150 and 800 miles were considered; these trips were then further filtered by keeping only those trips in the top decile (10%) of fares for each origin-destination mar- ket and by excluding any origin-destination markets where the corresponding number of average daily passengers was less than one. The logic behind the decile filter is that VLJ services are likely to be considered only by current commercial air busi- ness passengers who are already paying something close to full coach fares or business fares. However, the actual fare class data in the DB1B are not reliable, so the decile filter was used instead. The minimum market size restriction of one passenger per day should not significantly affect the re- sults since markets smaller than that collectively compose less than 6% of total trips. This resulted in estimates of cur- rent relevant commercial air passenger trips as shown in Table 10. The raw DB1B commercial air dataset is constructed on an airport-to-airport basis. As with the ETMS data, the DB1B passengers were ultimately distributed geographi- cally using Census population estimates to generate syn- thetic observations of commercial air travel at the Census tract level. Current Automobile Population The population of current automobile traffic that poten- tially could be affected by VLJ competition was derived from the 1995 ATS. This survey provided only limited geo- graphic information, including the MSA of the trip origin and/or destination if relevant. If the origin/destination was outside of an MSA, then only the State where the trip started or ended was identified. These automobile trips were ulti- mately assigned to specific VLJ catchment areas based on population and distance ratios. 15 Region of Departure Count Percent of Total 1028,098Central 1028,463Eastern 2567,881Great Lakes 411,860New England 822,791Northwestern Mountain 2671,249Southern 1028,051Southwestern 513,524Western Pacific 100271,917Total *Note: Figures and percentages may not add to totals due to rounding. Table 9. Relevant universe of air taxi flights for 2007.* Percent of TotalCountRegion of Departure 3479,019Central 183,292,341Eastern 152,796,151Great Lakes 4695,091New England 81,410,906Northwestern Mountain 193,558,592Southern 111,948,360Southwestern 234,342,707Western Pacific 10018,523,167Total *Note: Figures and percentages may not add to totals due to rounding. Table 10. Relevant universe of commercial air trips for 2007.*

Only domestic trips within the lower 48 states and between 150 and 400 miles (using the great-circle distance of the corresponding departure and arrival VLJ airports) were consid- ered; these trips were then further filtered by keeping only those trips taken by “high-income” business travelers, defined as individuals with annual incomes greater than $75,000 (in 1995 dollars). The logic behind imposing the 400-mile upper limit is that any automobile trips longer than that probably indicate that such travelers have particular reasons for selecting pri- vate surface travel (e.g., making multiple stops as a traveling salesman, etc.) and would not be good candidates for VLJ services. This resulted in estimates of relevant automobile trips as shown in Table 11. It is important to note that the mode choice decisions made by ATS automobile users reflect choices that were available to them in 1995. Consequently, the attributes of each mode alternative should reflect values relevant for that time period. As with the other datasets, the raw ATS data provide only limited information on the geographic distribution of pas- senger trips. Again, Census population estimates were used to generate synthetic observations of automobile travel at the Census tract level. Model Assumptions The data from the three sources described above—ETMS 2007 air taxi flights, DB1B 2007 commercial air trips, and ATS 1995 automobile trips—were combined into a single large dataset to reflect the potential universe of trips from which a new VLJ mode would attract customers. To convert flights into passenger trips for current air taxi users, an aver- age load factor of 70% was applied to the seat size of each rep- resentative aircraft selected. As noted earlier, the primary attributes used to distinguish one mode offering from another are cost and travel time. A representative aircraft was selected for each air taxi mode— piston, turboprop, and light jet. A review of aircraft-specific cost data indicated that real (inflation-adjusted) costs had not changed much between 1995 and 2007 for representative pis- ton or turboprop aircraft types, so the same costs were used for both automobile users and air taxi/commercial air users. Cost and travel time attributes were developed for each air- craft type based on data obtained from the Conklin & de Decker Aircraft Cost Evaluator database. Party size is an important factor in the statistical model be- cause the air taxi and automobile modes have capacity con- straints that could affect total costs depending on the party size (e.g., if more than one unit of the mode is required to accommodate the entire party). In addition, when the VLJ mode is added, it is important to consider the availability of per-seat service offers as this will also affect total costs. For existing air taxi users identified in the ETMS data, it was assumed that all passengers on board (estimated by applying a 70% load factor to the aircraft passenger seat size) were part of the same travelling party and that the mode choice decision was made for the group as a whole and not at the individual level. Current air taxi services are offered almost exclusively on a per-aircraft basis, so a reasonable assumption is that such flights mainly reflect demand by groups of travelers where it makes sense to rent the services of an entire aircraft for the group. For current (2007) commercial air users, the DB1B dataset does not include any information on party size. However, the ATS survey indicates that almost 75% of commercial air pas- sengers on business trips between 150 and 800 miles fly alone, so a party size of one was assumed for all current commercial air passengers. This is a somewhat optimistic assumption in favor of VLJ per-seat services because it does not consider that some commercial air trips are in fact taken by larger par- ties who might not realize the cost savings of per-seat services, but instead could realize savings as a group purchasing the services of more traditional air taxi services that are sold on a per-aircraft basis. Forcurrent (1995) automobile users, certain high-passenger- count records in the ATS dataset have large party sizes that are probably not typical; consequently, a maximum party size of two was used for automobile users in the ATS. This was also used as the capacity limit for the automobile alternative for current users of other modes. 16 Percent of TotalCountRegion of Departure Central Eastern Great Lakes New England Northwestern Mountain Southern Southwestern Western Pacific Total 5801,109 192,961,779 213,394,996 4684,720 5764,044 162,543,010 132,138,512 172,680,552 10015,968,722 Table 11. Relevant universe of automobile trips for 1995.

Explanatory Variables and Model Estimation Travelers choose from among five existing modes: auto- mobile, commercial air, air taxi piston, air taxi turboprop, and air taxi light jet.12 Each mode is characterized by a set of attributes that affect travelers’ choices. Aside from the stan- dard cost and time attributes applicable to each mode, a variable representing average time between departures was included for the commercial air mode to account for the effects of service availability of this mode. In addition, the num- ber of based prop or jet aircraft (averaged between the depar- ture and arrival airports) was calculated to serve as proxies for airport amenities that may affect the attractiveness of the three air taxi modes. Finally, personal income was also included as a user characteristic that may affect mode travel choice. A small number of different specifications were con- sidered before settling on the list of explanatory variables shown in Table 12. As indicated in the table, the effect of the constant term, cost, and income variables was allowed to vary by mode; for example, the effect of a change in the cost of driving for those passengers selecting automobile was allowed to be different from the effect of a change in the cost of commercial air for those selecting that mode. Additional details about the assump- tions and calculations employed are shown in Table 13. Although cost and travel time are typically primary factors in any mode choice analysis, it is important to recognize that other more difficult-to-measure attributes such as travel conve- nience, reliability, and unforeseen congestion may also be significant factors in choosing among available modes. These are accounted for indirectly in the model by includ- ing mode-specific constant terms that, in principle, pick up average unobserved effects of each individual mode.13 The model is estimated with a “multinomial logit” specifi- cation, and attempts to ascribe coefficient values to the attri- butes that best fit the observed choices made in the datasets (ETMS, DB1B, and ATS) described above. Because there are literally hundreds of thousands of observations derived from these datasets, it was not feasible to incorporate all of them into the estimation process. Instead, a sample of observations was drawn from each dataset in such a way as to ensure that 17 Variable Automobile Commercial Air Air Taxi Piston Air Taxi Prop Air Taxi Light Jet Constant Term (Mode-specific) Cost (Mode-specific) Total Travel Time Avg Time Between Departures (Commercial Air) Income (Mode-specific) log(Based GA Piston + Prop Aircraft) log(Based GA Jet Aircraft) Table 12. List of explanatory variables used in air taxi analysis. 12The automobile mode is excluded from the choice set when the auto alternative would involve driving more than 400 miles. Cost: AAA total driving cost per mile for intermediate car (2006) x mileage computed from Microsoft MapPoint software. Travel Time: Drive times from Microsoft MapPoint software (+ periodic rest stops for trips greater than 3 hours). Cost: Top decile DB1B FY2007 market-specific fare from nearest commercial airports + access/egress cost to airports from Census locations. Travel Time: Weighted-average travel time for market-specific non-stop/one-stop/two- stop services (from May 2007 OAG schedule) + access/egress times to airports + airport terminal wait times. Average Time Between Departures: Market-specific estimate based on number of daily service offers assuming 16-hour day. Cost: See Model Assumptions section. Travel Time: Based on average aircraft speeds + access/egress times to airports. Based Aircraft: The number of based aircraft was measured in logs, allowing the positive impact of this “attractiveness” proxy to lessen at the margin. Automobile Commercial Air Air Taxi Table 13. Explanatory variable calculations. 13As with the income variable whose value for a given individual does not vary across modes, it is necessary to “normalize” on one of the modes as the baseline (statistically it does not matter which one). Then, the income or constant term effects are interpreted relative to the baseline mode.

18 Variable Coefficient Estimate t-statistic 897.51044.7elibomotuA-tnatsnoC Constant - Commercial Air -5.3354 -3.895 885.28124.3notsiPixaTriA-tnatsnoC 872.0-1416.0-porPixaTriA-tnatsnoC 424.12-3755.0-elibomotuA-tsoC 163.11-7210.0-riAlaicremmoC-tsoC 731.51-7310.0-notsiPixaTriA-tsoC 224.7-8700.0-porPixaTriA-tsoC 079.11-1700.0-teJthgiLixaTriA-tsoC 260.6-3900.0-emiTlevarTlatoT Avg Time Between Departures - Commercial Air -0.0049 -10.923 043.0-8500.0-elibomotuA-emocnI 223.54490.0riAlaicremmoC-emocnI 102.2-5040.0-notsiPixaTriA-emocnI 691.1-4530.0-porPixaTriA-emocnI Based GA Piston + Prop Aircraft 0.9307 3.664 290.57358.0tfarcriAteJAGdesaB Table 14. Logit model coefficient estimates. Variable Automobile Commercial Air Air Taxi Piston Air Taxi Prop Air Taxi Light Jet 71.3-15.5-25.2-81.0-07.0-)ecirP(tsoC 05.0-01.1-84.0-71.0-23.0-emiTlevarT Income (relative to AT light jet) -0.03 0.35 -0.90 -1.94 NA 90.0-serutrapeDneewteBemiTgvA 98.345.1porp/notsipAG-tfarcriAdesaB Based Aircraft - GA jet 1.17 Table 15. Direct elasticity estimates. the overall sample is representative of the total population shares of each mode derived from the datasets. These existing mode shares are as follows: • Automobile: 44.9%; • Commercial Air: 52.1%; • Air Taxi, Piston: 2.0%; • Air Taxi, Turboprop: 0.2%; and • Air Taxi, Light Jet: 0.8%. Statistical Results The coefficient estimates and statistical significance indica- tors are shown in Table 14. T-statistics greater than about 2.0 in absolute value indicate statistical significance at the 95% confidence level. As can be seen, most coefficients are statisti- cally significant, with the exception of two alternative-specific income variables and the constant for the air taxi piston mode. All variables listed in Table 14 (even those with insignificant coefficient estimates) were included when making projec- tions for future years. There is no easy straightforward interpretation for the co- efficients themselves; not even the sign of the coefficients necessarily indicates the direction of effects. A more meaning- ful interpretation can be gained by computing so-called “direct mode elasticities,” which reflect how a 1% change in the value of a particular attribute for a particular mode will affect the likelihood of selecting that mode. For example, a price elastic- ity of −2 for the commercial air mode means that a 1% increase in the price of commercial air would lead to a 2% decline in the probability of selecting that mode. Direct elasticities for all of the explanatory variables in the model are shown in Table 15. For the most part, the elasticity estimates have the expected sign. Increases in price, travel time, and average time between flights all lead to decreases in the probability of selecting the associated mode. Increases in the number of based aircraft (piston/props or jets) at GA airports are associated with in- creases in the probability of selecting one of the air taxi modes. If in fact these based aircraft counts are reasonable proxies for airport amenities such as hangar facilities, ground transporta- tion services, precision approaches, and so forth, the relatively large values of the elasticities suggest that the ability of airports to provide such amenities may in fact lead to significant new air taxi traffic. Somewhat surprisingly, the elasticity estimates also indi- cate that a given price change for the air taxi modes will elicit

larger changes in the likelihood of selecting those modes than comparable changes in the automobile or commercial air modes—in other words, air taxi users are relatively more price- elastic than automobile or commercial air users. While this may initially seem counter-intuitive, one must recognize that the elasticities are “point estimates” that reflect current observed price levels and mode shares. For example, decreasing an air taxi jet price of $1,200 by 1% would lead to a 3.15% increase in the likelihood of selecting that mode (which currently has less than 1% market share); this compares with decreasing the corre- sponding commercial air price of, say, $300 by 1% leading to a much smaller (0.18%) increase in the likelihood of selecting that mode (which currently has a 52% market share). The interpretation of the effects of income on the choice de- cision is fairly difficult due to the choice-specific specification for income; this means that the direction of effect is relative to the normalized air taxi jet alternative. However, there is only one income elasticity indicating a relatively large nominal effect (−1.86 for air taxi turboprop); in addition, there are only a small number of travelers who selected air taxi turboprop. Consequently, it is believed that income effects in the model are relatively unimportant and do not have a large effect on the overall results. To assess the overall fit of the model to the data, a “pseudo- R2” statistic was computed, which is somewhat analogous to the standard R2 statistic often reported in linear regressions. Using a scale of 0 to 1, the statistic is an indicator of how well the model fits the observed data. The pseudo-R2 estimate is 0.675, which most analysts would consider quite good for a multinomial logit model. Another useful measure is to compute the implied values of time for each mode. This is accomplished by dividing the com- mon time coefficient by the alternative-specific cost coefficients. Table 16 shows the estimated value of time; these are consistent with the expectation that travelers selecting the air taxi modes have higher values of time than those selecting the automobile or commercial air modes. These estimates are also generally consistent with FAA guid- ance on value-of-time estimates for business travelers using commercial air as published in its latest “Economic Values for FAA Investment and Regulatory Decisions.” They are some- what above the FAA estimates published for GA. Baseline Forecast Assumptions With the coefficient estimates in hand, the overall number of trips from the ATS dataset (which was based on trips taken in 1995) was scaled up to account for growth in the overall magnitude of travel between 1995 and 2007. All 1995 data val- ues used for automobile users were updated to 2007 values, and increases in the time spent in terminal areas at commercial airports (due to increased security measures since 2001) were accounted for in the commercial air mode. Then the revised population total (including commercial air and air taxi) was combined with the coefficient estimates from the model popu- lations to generate estimates of trips by mode for the baseline year 2007.14 Each trip is tied to specific GA and commercial air- ports that would be relevant when making the mode choice decision, so baseline estimates of air taxi activity at each of the 1,842 “VLJ-ready” airports for 2007 can be produced. These estimates will be provided electronically as the appendix to the second volume of this report, ACRP Report 17: Airports and the Newest Generation of General Aviation Aircraft, Volume 2: Guidebook. When reviewing these estimates, it is important to recognize that they account only for air taxi activity. Outlook for Air Taxi Services Utilizing Small Next-Generation Aircraft The next step is to add in the new VLJ mode and prepare pro- jections of trips by mode for the forecast years 2012 and 2017. Although the 2007 baseline estimates themselves do not include any VLJ activity by startup air taxi operators, there are in fact a number of existing startups that have already begun to use VLJs or other small next-generation aircraft for air taxi–type services. There are a variety of business models being tried. The traditional “air charter” model typically involves ex- clusive rental of an entire aircraft for a fixed hourly rate that covers the cost of the aircraft, including pilot salaries and fuel costs. Additional costs can include taxes, repositioning fees, and overnight/waiting fees. Even if a return trip is not needed, there will likely be a charge for the cost of repositioning the aircraft to its home (or other) location, and there may also be a daily minimum charge. The “air taxi” on-demand model involves rental of an entire aircraft for a fixed hourly rate, but no charges for repositioning or overnight/waiting times; all costs are built into the hourly rate. If there is a return trip, it may be on a different aircraft (or even provided by a different company). This type of service may only be available between certain specified airports. The newest business model attempted by some operators is the “per-seat on-demand” model, which is somewhat similar 19 Mode Estimated Value of Time ($/hr) 10.02Automobile 44.04Commercial Air 40.78Air Taxi Piston 72.53Air Taxi Prop 78.97Air Taxi Light Jet Table 16. Implied value of time estimates. 14Where necessary, small adjustments were made to the alternative- specific constants to obtain passenger trip estimates for each mode that were consistent with the observed baseline of 2007 trips.

to buying a ticket for an individual seat from a commercial air- line, but there is no fixed flight schedule. The price may depend on the number of other passengers actually on board the flight, or it could be pre-determined based on the average number of passengers the operator expects. There may also be other pric- ing structure variations such as offering a lower pre-determined price if the passenger is willing to travel any time within some pre-defined time window. These service models have initially sprung up primarily in the Southeast and Northeast sections of the country, although similar operations have been planned (and some are already operating) in the Midwest and on the West Coast. Some are using small, efficient, next-generation piston-based aircraft while others have taken deliveries of small numbers of VLJs. Despite the 2008 bankruptcy of DayJet, which was a pioneer in the development and offering of per-seat on-demand ser- vices, current plans from other startups appear to indicate that many still believe that the concept of the per-seat business model can be viable. It is useful to consider these operators and their service plans in order to make reasonable assumptions about how such service offers may spread throughout the United States for the 5- and 10-year forecasts. Careful consideration has been given to the pricing approaches, geographic location, and fleet types of such operators in the consideration of how to model a new “VLJ” mode that will be added to the choices available to consumers. VLJ Mode Attributes The new VLJ mode was represented by taking an average of the estimated operating cost and capacity attributes of sev- eral models. In addition, the VLJ mode was assumed to in- herit the same alternative-specific constant, cost, and airport jet presence coefficients as the air taxi light jet mode. As with the other air taxi services, cost to the traveler was estimated using a 75% markup to the aircraft operating costs estimated in the Conklin & de Decker data. Spread and Distribution of VLJ Per-Seat versus Traditional Charter Services It is believed that a per-seat pricing approach still could be the foundation for a successful business model for VLJ air taxi services, although the prices may need to be somewhat higher than initially estimated or selection and utilization of aircraft type may need to be altered. To assess the potential impact of the per-seat approach to air taxi services, an analysis of actual fares offered for per-seat services in 2008 was undertaken. On average, it was found that per-seat prices with a 3- to 4-h departure window were priced at about a 40% discount to those with a 1-h departure window in the same market. The latter was assumed to be similar to the prices that would be charged for renting the en- tire aircraft (along the lines of a more traditional charter or air taxi service). For the 2012 baseline forecast, it was assumed that the VLJ mode would be offered with per-seat pricing (and the asso- ciated 40% discount) in the following FAA regions: South- east, Southwest, and Western Pacific. Notionally, these are consistent with prospects for viable per-seat operations that were found via interviews and discussions with various in- dustry participants, but it should be noted that the implicit as- sumption is that all VLJ service offers in these regions are via the per-seat model and none are with the traditional charter model. It was assumed that per-seat pricing would also entail wait time equivalent to commercial service that offered approxi- mately four flights per day; this was valued using the same time-between-departure coefficient that applies to commer- cial air travel. Thus, the benefit of a lower price is partially off- set by the wait time that one must incur relative to traditional charter service. In all other regions, it was assumed that the VLJ mode would be available by 2012, but only via traditional charter (per aircraft) with the associated higher cost. For the 2017 forecast, it was assumed that per-seat pricing would spread to the Great Lakes region. Overall, in light of the recent down- turn in the economy and bankruptcies in the market, these baseline assumptions of VLJ availability by 2012 and 2017 may be relatively optimistic. Spread and Distribution of Low-Cost Piston Services The analysis also takes account of the likelihood that low- cost piston air taxi services using efficient new-generation aircraft will expand beyond current service areas centered in the Southeast region of the United States. For the baseline forecast, it was assumed that low-cost service would replace higher-cost traditional piston service in the rest of the South- east, Southwest, Western Pacific, and Great Lakes by 2012. By 2017, it is assumed that traditional piston services are re- placed by low-cost alternatives entirely throughout the United States. Other Modes With the exception of the impact of fuel prices on total cost (see below), the attributes of the remaining modes— automobile, commercial air, and air taxi turboprop and light jet—were assumed to remain constant for all geographic regions through the forecast years 2012 and 2017. 20

Impact of Increasing Fuel Prices The baseline model projections for 2007 were estimated using costs that were based on late 2006–2007 fuel prices. The operating costs for the air taxi modes were based on avi- ation gasoline (or “avgas,” which is used in piston aircraft) and jet fuel prices of $2.45 per gallon while automobile op- erating costs were based on a pump price of $2.94 per gallon. Given the volatility in the price of oil, it is difficult to project future costs and prices for the 2012 and 2017 forecasts. Since late 2007, the world price of oil swung wildly upward during the first half of 2008 and then plummeted dramatically in the second half. For projection purposes in 2012 and 2017, costs for the air taxi modes were re-computed assuming an average price of $3.35 per gallon for both avgas and jet fuel. This is equal to the U.S. Department of Energy’s (DOE’s) jet fuel price average for the first eight months of 2008 and equates to an overall oper- ating cost increase from 2007 of about 3.8% for the low-cost piston mode and about 6.5%–7.0% for the turboprop and light jet modes. For the automobile mode, the average cost per mile was re-computed assuming an average pump price of $3.61 per gallon, again based on the DOE’s estimates for the first eight months of 2008; this equates to an overall operating cost increase from 2007 of 8.8%. Changes in oil prices would also be expected to influence commercial air fares. For the commercial air mode, the lat- est available fare data from the DB1B ticket sample were obtained for the second quarter of 2008; a comparison with FY2007 fares showed an average fare increase of 6.0%, and this percentage increase was assumed to apply for the 2012 and 2017 projection years. Overall Travel Growth As described earlier, the baseline forecast assumed a constant per-capita trip rate for the domestic United States, and overall growth in travel trips is therefore proportional to population growth. Census-specific growth projections were used from the AGS dataset; across the entire sample, the corresponding pop- ulation is projected to grow by about 1.1% annually from 2007 to 2012 and by about 1.4% annually from 2007 to 2012. Projected Mode Shares and Trip Totals Table 17 shows the projections of mode shares and trip to- tals given all of the inputs and assumptions described above. The forecast projects that VLJs may capture a small but sig- nificant percentage of the relevant market over the next 5 to 10 years. This is due primarily to the assumption that low-cost per-seat services will become available in several regions in the country. Similarly, the assumption that low-cost piston ser- vices will spread leads to significant gains for that mode as well. These gains in market share come largely at the expense of the automobile mode and, to a lesser extent, of commercial air travel. Without reading too much into the data, these results are generally consistent with the observation that automobile traffic in particular may be quite responsive to large increases in fuel prices such as those that occurred in mid-2008. 21 detcejorP7102detcejorP2102detcejorP7002lautcA7002edoM 846,71446,61835,81096,81elibomotuA 021,02060,02126,81325,81riAlaicremmoC 505,1862,1927596notsiPixaTriA 37143160137porPixaTriA 963792562872teJthgiLixaTriA 914,3969,100JLVixaTriA 532,34273,04952,83952,83LATOT detcejorP7102detcejorP2102detcejorP7002lautcA7002edoM %8.04%2.14%5.84%9.84otuA %5.64%7.94%7.84%4.84riAlaicremmoC %5.3%1.3%9.1%8.1notsiPixaTriA %4.0%3.0%3.0%2.0porPixaTriA %9.0%7.0%7.0%7.0teJthgiLixaTriA %9.7%9.4%0.0%0.0JLVixaTriA %0.001%0.001%0.001%0.001LATOT Annual Trips (000) Market Shares *Note: Figures and percentages may not add to totals due to rounding. Table 17. Estimated air taxi annual trips and market shares.*

Baseline Air Taxi Fleet Forecast In order to translate the traffic projections into air taxi fleet forecasts, assumptions must be made regarding air- craft use and load factors. For the baseline forecasts, it is assumed that the air taxi modes use their fleets at the rate of two flights per day for piston, turboprop, and light jet and three flights per day for VLJs (due to the spread of per- seat services). An average passenger load factor of approx- imately 70% is also assumed. With an average flight time of about 1.3 h, this works out to approximately 1,200 h of uti- lization per year for VLJs and 800 h for the other aircraft types. This is well above current utilization rates for small GA aircraft (which are more on the order of a few hundred hours per year), but still only a fraction of the utilization rates typical of large commercial aircraft. The required fleets to provide the projected trips at these rates are shown in Table 18.15 It is important to recognize that the actual air taxi fleet pro- jection levels depend heavily on a number of basic assump- tions, the most prominent of which are • Definitions of the relevant universe for the automobile and commercial air travel markets; • “Full price of travel” estimates of the various modes, which depend on (among other things) uncertain estimates of the unit costs of providing traditional charter operations, wait and/or delay times associated with commercial air travel, and road congestion associated with automobile travel; • Actual availability of new “per-seat” VLJ services and/or low-cost piston services; and • Perceived similarities or differences between new services and traditional charter services. Using different assumptions for any of these factors could have significant impacts on the estimated results. Operational Impacts on Airports Given the baseline projections, the trip estimates are all tied to specific locations and airports, so the total number of associated air taxi operations on an airport-specific basis can be aggregated. The projected operational increases for 2017 compared with the TAF 2007 total operation estimates are summarized in Table 19; the average operational increase by 2017 relative to the 2007 baseline is on the order of 6%. 22 Mode Avg Passengers per Flight @70% LF 2007 Projected 2012 Projected 2017 Projected Net Increase 2007-2012 Net Increase 2007-2017 Air Taxi Piston 2.1 556 968 1,149 411 593 Air Taxi Prop 3.5 49 61 79 13 31 Air Taxi Light Jet 4.2 101 113 141 12 39 Air Taxi VLJ 2.8 0 751 1,305 751 1,305 769,1881,1376,2498,1607LATOT *Note: Figures and percentages may not add to totals due to rounding. Table 18. Air taxi fleet forecast.* Region TAF 2007 Total Operations 2017 Incremental Air Taxi Operations Air Taxi % 2007 Operations 9.4157,441229,379,2lartneC 2.5274,814880,200,8nretsaE 0.6353,947508,655,21sekaLtaerG 6.2839,18221,601,3dnalgnEweN 0.3373,252379,193,8niatnuoMnretsewhtroN 1.5907,419759,029,71nrehtuoS 4.6347,256471,951,01nretsewhtuoS 3.8441,639660,813,11cificaPnretseW 6.5384,051,4701,924,47latoT Table 19. Estimated incremental air taxi operations by region. 15The 2007 projections from the model are well below the current aggre- gate air taxi fleet shown earlier in Table 2. The two really cannot be compared for a variety of reasons, including that (1) the FAA air taxi def- initions from Table 2 are quite different and cover a much larger portion of overall flight activity than the actual usage assignments used here based on ETMS activity and (2) the utilization rates used in this analysis are much higher than the historical averages for piston, turboprop, and jet categories shown in Table 3, which are confounded because they reflect combined activity of aircraft across different usage categories.

Obviously the results may vary significantly for specific facilities; airport-specific air taxi estimates for all 1,842 air- ports included in the study are provided electronically in the appendix for Volume 2 of this report. The five airports in each FAA region with the largest increases in projected activity by 2017 from VLJ operations are shown in Table 20; corresponding forecasts that account for increased activity by all air taxi modes including pistons, turboprops, and light jets are shown in Table 21. In some cases, the large in- creases shown are due primarily to the VLJ or low-cost piston modes capturing significant shares of popular auto- mobile traffic corridors such as between Southern California and Las Vegas and the Texas triangle connecting Dallas, Houston, and San Antonio. 23 etatSytiCytilicaFdicoLnoigeR FAA Itinerant + Local Ops 2007 Added VLJ Air Taxi Ops by 2017 038,21483,641OMSt. LouisSpirit of St. LouisSUS MKC Charles B. Wheeler Downtown Kansas City MO 95,438 8,283 596,6112,601AIsenioMseDltnIsenioMseDMSD OJC Johnson County Executive Olathe KS 70,438 3,058 SGF Springfield-Branson National Springfield MO 74,504 2,198 HPN Westchester County White Plains NY 202,572 25,259 276,32391,202JNorobreteTorobreteTBET SYR Syracuse Hancock Intl Syracuse NY 107,749 16,506 431,51314,93YNsllaFaragaiNltnIsllaFaragaiNGAI AGC Allegheny County Pittsburgh PA 82,185 13,858 PWK Chicago Executive Chicago/Prospect Heights IL 118,496 42,987 LUK Cincinnati Muni Airport Lunken Field Cincinnati OH 72,717 30,942 405,31869,48IMtiorteDnuRwolliWPIY DET Coleman A. Young Muni Detroit MI 77,571 12,033 ATW Outagamie County Rgnl Appleton WI 46,440 11,897 OWD Norwood Memorial Norwood MA 84,784 12,873 BED Laurence G. Hanscom Fld Bedford MA 169,471 4,775 455,3153,96AMylreveBinuMylreveBYVB 339,1732,401HNauhsaNdleiFerioBHSA OXC Waterbury-Oxford Oxford CT 60,829 1,708 155,91959,923OCrevneDlainnetneCAPA BJC Rocky Mountain Metropolitan Denver CO 167,968 6,188 DRO Durango-La Plata County Durango CO 57,123 4,845 655,4700,47OCnoitcnuJdnarGdleiFreklaWTJG HIO Portland-Hillsboro Portland OR 224,461 4,534 PDK Dekalb-Peachtree Atlanta GA 223,399 35,184 FTY Fulton County Airport-Brown Field Atlanta GA 122,196 29,093 LZU Gwinnett County - Briscoe Field Lawrenceville GA 85,686 26,682 712,710CNlliHlepahCsmailliWecaroHXGI RYY Cobb County-Mc Collum Field Atlanta GA 110,069 17,018 071,29338,131XTsallaDnosiddASDA 264,62136,841XToinotnAnaSinuMnosnitSFSS 813,32835,68XTnotsuoHlngRdnaLraguSRGS 753,81479,441XTnotsuoHdleiFnotgnillEDFE HYI San Marcos Muni San Marcos TX 120,420 12,984 556,29042,912VNsageVsaLsageVsaLhtroNTGV CRQ Mc Clellan-Palomar Carlsbad CA 215,859 28,667 MYF Montgomery Field San Diego CA 223,410 25,674 HND Henderson Executive Las Vegas NV 67,482 22,006 FAT Fresno Yosemite Intl Fresno CA 156,648 18,705 Northwestern Mountain Southern Southwestern Western Pacific Central Eastern Great Lakes New England Table 20. Added VLJ air taxi operations by 2017—top five airports by region.

24 etatSytiCytilicaFdicoLnoigeR FAA Itinerant + Local Ops 2007 Added Total Air Taxi Ops by 2017 798,61483,641OMSt. LouisSpirit of St. LouisSUS MKC Charles B. Wheeler Downtown Kansas City MO 95,438 14,359 487,8112,601AIsenioMseDltnIsenioMseDMSD SGF Springfield-Branson National Springfield MO 74,504 5,970 OJC Johnson County Executive Olathe KS 70,438 5,784 HPN Westchester County White Plains NY 202,572 31,561 743,92391,202JNorobreteTorobreteTBET SYR Syracuse Hancock Intl Syracuse NY 107,749 23,103 AGC Allegheny County Pittsburgh PA 82,185 18,691 HEF Manassas Rgnl/Harry P. Davis Field Manassas VA 110,132 18,275 PWK Chicago Executive Chicago/Prospect Heights IL 118,496 51,870 LUK Cincinnati Muni Airport Lunken Field Cincinnati OH 72,717 42,304 PTK Oakland County Intl Pontiac MI 209,198 17,808 037,71869,48IMtiorteDnuRwolliWPIY DET Coleman A. Young Muni Detroit MI 77,571 17,449 OWD Norwood Memorial Norwood MA 84,784 15,697 BED Laurence G. Hanscom Fld Bedford MA 169,471 7,010 484,4153,96AMylreveBinuMylreveBYVB 769,2732,401HNauhsaNdleiFerioBHSA MVY Marthas Vineyard Vineyard Haven MA 52,060 2,706 504,03959,923OCrevneDlainnetneCAPA PAE Snohomish County (Paine Fld) Everett WA 131,836 13,455 BJC Rocky Mountain Metropolitan Denver CO 167,968 10,957 HIO Portland-Hillsboro Portland OR 224,461 10,177 TTD Portland-Troutdale Portland OR 86,721 10,154 PDK Dekalb-Peachtree Atlanta GA 223,399 47,433 FTY Fulton County Airport-Brown Field Atlanta GA 122,196 34,773 LZU Gwinnett County - Briscoe Field Lawrenceville GA 85,686 34,213 RYY Cobb County-Mc Collum Field Atlanta GA 110,069 20,955 772,02437,151LFodnalrOevitucexELRO 889,111338,131XTsallaDnosiddASDA 656,73136,841XToinotnAnaSinuMnosnitSFSS 741,13835,68XTnotsuoHlngRdnaLraguSRGS 690,52479,441XTnotsuoHdleiFnotgnillEDFE HYI San Marcos Muni San Marcos TX 120,420 18,187 991,052042,912VNsageVsaLsageVsaLhtroNTGV MYF Montgomery Field San Diego CA 223,410 78,118 CRQ Mc Clellan-Palomar Carlsbad CA 215,859 60,820 440,65243,592ACnojaClE/ogeiDnaSdleiFeipselliGEES HND Henderson Executive Las Vegas NV 67,482 47,108 Northwestern Mountain Southern Southwestern Western Pacific Central Eastern Great Lakes New England Table 21. Added total air taxi operations by 2017—top five airports by region.

Next: Chapter 4 - Conclusions »
Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB’s Airport Cooperative Research Program (ACRP) Report 17: Airports and the Newest Generation of General Aviation Aircraft, Volume 1: Forecast explores a forecast of anticipated fleet activity associated with the newest generation of general aviation aircraft for 5- and 10-year outlooks. ACRP Report 17, Volume 2 is a guidebook designed to help airport operators assess the practical requirements and innovative approaches that may be needed to accommodate these new aircraft.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!