Journal of Aviation/Aerospace Education & Research

Volume 2 Number 2 JAAER Winter 1992 Article 6

Winter 1992

A Method for Identifying General Aviation Airports that are Candidates for Runway Extensions: A Planning Model for State Aviation Systems

Randall G. Holcombe

Henry B. Burdg

Follow this and additional works at: https://commons.erau.edu/jaaer

Scholarly Commons Citation Holcombe, R. G., & Burdg, H. B. (1992). A Method for Identifying General Aviation Airports that are Candidates for Runway Extensions: A Planning Model for State Aviation Systems. Journal of Aviation/ Aerospace Education & Research, 2(2). https://doi.org/10.15394/jaaer.1992.1068

This Article is brought to you for free and open access by the Journals at Scholarly Commons. It has been accepted for inclusion in Journal of Aviation/Aerospace Education & Research by an authorized administrator of Scholarly Commons. For more information, please contact [email protected]. Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi A METHOD FOR IDENTIFYING GENERAL AVIATION AIRPORTS THAT ARE CANDIDATES FOR RUNWAY EXTENSIONS: A PLANNING MODEL FOR STATE AVIATION SYSTEMS

Randall G. Holcombe and Henry B. Burdg

ABSTRACT

One of the most important characteristics of an airport is the length of its longest runway: that length determines the types ofaircraft that can use the airport andprovides a margin ofsafety for users. Arunway extension, therefore, would enhance the utility ofmany airports. But resources are scarce, and iffunds are to be allocated at the state level, state officials need a method for determining which airports could best use a longer runway. Consideration of demographic factors enhances the traditional engineering analysis approach to runway evaluation. This combined approach is more consistent with the goals of comprehensive airport planning. This paper describes a model that uses regression analysis to compare airports in Alabama, taking into account a number of different demographic factors and airport related factors. A linearregression model was used to evaluate howan airport's runway length compared to others around the state with similar characteristics. The residuals from the regressions were used to identify those airports with relatively short runways, considering their other characteristics. The regression analysis identified 22 of the 106 public-use airports in Alabama as having runways substantially shorter than their other characteristics would predict.

One of the most important 1988) estimated that $285 million categories: demographic factors characteristics of an airport is the (1988-1989) in state funds were (Federal Aviation Administration length of its longest runway, spent on state airport develop­ [FAA], 1972) and airport related because it will determine the ment projects. In many situations factors (Horonjeff & McKelvy, types of aircraft that can use the it appears that these funds are 1983). In this paper 10 demo­ airport and because longer run­ politically directed rather than graphic factors are considered. ways provide a margin of safety proactively planned. including the population of the for any aircraft using the airport Resources are scarce and if county where the airport is (Ashford & Wright. 1979). A short state airport development funds located. population in the runway limits the usefulness of are to be allocated effectively. immediate vicinity of the airport, an airport and restricts the ability state officials need a method for population growth in the area, to accommodate the current evaluating candidate airportsthat population density, and the corporate fleet. Many of the could best use longer runways. number of nearby businesses, general aviation airports were This paper describes a model employees, and payroll. The designed and constructed 35 or that uses regression analysis to paper also considers 23 airport more years ago. meeting the compare airports in Alabama. factors, including the number of aircraft requirements of that day. taking into account a number of based aircraft and operations at For the most part the smaller different factors. The work the airport, services offered. and community airport has not described here is included in a the locations of other airports. In developed as fast as the larger study of general aviation certain geographical settings. it population it is to serve. A airports in Alabama conducted is important to consider the runway extension, therefore. for the State of Alabama. Depart­ variation in airport elevations. would enhance the utility of ment of Aeronautics. March runway gradient, and design many airports. 1987. With the results from this reference temperatures. These There are some 5,598 general model. state officials can better factors do affect runway length. aviation airports and 568 general evaluate proposals for runway In Alabama the variation is repre­ aviation and commercial airports extensions. sented by a narrow range and in the United States. A recent Factors determining the thus excluded from the analysis. study (National Association of appropriate runway length can For exar.lple. the design refer­ State Aviation Officials [NASAO], be divided into two general ence temperatures range from

Published by Scholarly14 Commons, 1992 JAAER, Winter 1992 1 Journal of Aviation/Aerospace Education & Research, Vol. 2, No. 2 [1992], Art. 6 Runway Extension Planning Model

89-94 degrees F. Another con­ operate from a runway 2,000 feet conjunction with the runway sideration can be the initial long, but a heavy single-engine engineering specifications design function of the airport. or multi-engine craft needs a required by the airport'S past For example, a few Alabama air­ runway of at least 3,000 feet. designated Critical Aircraft. ports were initially built as mili­ Although small jets can take off Ideally, a state would compare tary airfields and converted to and land on 4,000 foot runways, its airports, taking into account public general aviation facilities. larger jets require 5,000 feet or the differences in their environ­ These tend to have longer run­ more. Runways of 10,000 feet ments, and compile a list of ways than airports specifically are common at the major com­ those that could benefit most designed and constructed for mercial service airports (FAA, from runway extensions. Many general aviation use. Additional 1983). airports could benefit from longer variables can be added to the These runway lengths are runways, but this paper presents regression models as needed to approximations. Under many the results of a regression model accommodate a region's unique circumstances, the aircraft that identifies those with features. described above could operate relatively short runways, taking A linear regression model was from shorter runways, even other factors into account. The used to compare the lengths of though a margin of safety would remainder of this paper explains the runways at general aviation be lost. An experienced pilot, for howthe analysis was undertaken airports around the state. Several example, would have no trouble for Alabama's airports. regression equations were esti­ landing a typical civilian training METHOD· DATA USED mated with runway length as the aircraft in 2,000 feet of runway, IN THE STUDY dependent variable. One regres­ but the runway would probably Alabama has 106 public-use sion used all 32 independent be too short for a student pilot or airports, as identified by Federal variables; others used only a pilot who flies infrequently. The Aviation Administration (FAA) demographic factors, airport trade-off is clear. A longer run­ 1986 records. The FAA maintains data, or number of operations at way will make an airport safer a Form 5010-1, Airport Master the airport. The residuals from and more useful, but runway Record, for each federally the regressions were used to extensions are expensive. If registered airport in the United identify those airports with runway extension projects are States. This study analyzes all relatively short runways, eligible for state airport develop­ public-use airports in the state, consideringtheirothercharacter­ ment funds, state airport officials including those that are privately istics. The regression analysis need a means of identifying owned, for which the FAA main­ identified 22 of the 106 public­ those airports that would benefit tains a 5010-1 form. The study use airports in Alabama as most from longer runways. excludes military airports, having runways substantially Examining length alone is heliports, and seaplane bases. shorter than their other charac­ insufficient: a 3,500-foot runway Table 1 presents a list of the teristics would predict. adequate for an infrequently variables collected for each After the regression analysis is used airport catering to single­ airport and used in the analysis. explained, the airports are engine traffic, would be a signi­ The first 10 are demographic examined in detail to determine ficant constraint for a community variables that characterize the why their runways are shorter that has the potential for many population that would use the than expected and to explain operations involving larger airport. how the regression results can aircraft, including jets. Any The variable County is the be applied to make a recommen­ determination ofthe adequacy of population ofthe county in which dation to lengthen a runway. a runway must take into account the airport is located. Most RUNWAY LENGTH AND the economic and demographic counties in Alabama have only AIRPORT UTILITY characteristics of the community one airport, but a significant The usefulness of an airport is or region served by the airport number (22 of 67) have more determined in large part by the and the use to which the airport than one. ~ is the population length of its longest runway. A is put (FAA, 1975). These factors of the primary city served by the light, single-engine airplane can are taken into account in airport. These data were https://commons.erau.edu/jaaer/vol2/iss2/6 DOI: https://doi.org/10.15394/jaaer.1992.1068JAAER, Winter 1992 ·15 2 Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi Runway Extension Planning Model

obtained from the 1980 Census. served by the airport and the aviation airport. These variables PopSrv is the population served type of business activity (Appch and LOC) are included by the airport, which was deter­ supported by that population. A because an airport equipped for mined by examining a map and larger population with more operations in poor weather is a matching small districts within a income would warrant better better candidate for a longer county to the nearest airport. facilities to support aviation runway, other things being The percent county population activity, so any study of runway equal. Information for these growth ~ was also taken length should take these two variables is from NOS Approach from the 1980 U.S. Census. factors into account. Plates-the set of maps a pilot The variable manufacturing Variable 11 is RwyLen (runway would use to fly the approach. employment (MfgEmp) was length), which is the dependent The final two variables indi­ obtained from the Alabama variable. in the stUdy. cate conditions that would tend 1985-86 Directory of Mining and Variables 12 through 17, taken to lessen the demand for a Manufacturing. Every manufac­ from the airport's FAA 5010-1 longer runway. The first (Nearby) turer was associated with the forms, describe the services indicates that another airport is nearest airport, and the number available. An airport supporting nearby; the second (SecArot), representing employment was air charter services, instruction, that the facility being analyzed is the sum of all employees in firms etc., will utilize a longer runway a secondary airport for a nearby with over 100 employees. Firms better than one without these primary airport. with more than 100 employees services. Variables indicating a certain are more likely than smaller firms The next-four variables, 18-21, condition are binary variables, to use aircraft in the course of list by type the aircraft based at taking on the value of -1- if the their business. The number of the airport. Again, more aircraft condition exists and -0- other­ employees thus determined based at the field would indicate wise. The binary variables are provides one indication of manu­ the demand for a longer runway. 12-17 and 28-33 in Table 1. For facturing activity in an area. Multiengine and jet aircraft example, if flight instruction is The next three variables, warrant greater runway length, available at an airport, the value county employment (CouEmp), so they are listed separately. of that variable would be given 1; the number of businesses Variables 22-27 give the num­ if flight instruction is unavailable, (#Bus), and county payroll ber of aircraft operations at a the value of the variable would (Payroll), are additional indicators field, broken down by type of beO. of business activity near the flight. The next two variables Each variable should be taken airport. These were taken from indicate whether the field has into account in determining the County Business Patterns, 1983, lighting for night operations and appropriate runway length for an Alabama. whether a control tower is airport. In the regression analysis Manufacturing percent located on the field. Variables these various factors interact to (MfgPet) is an indication of the 18-29 are also from the FAA predid the airport's runway percentage of county employ­ 5010-1 forms. length, given its other character­ ment that is made up of manu­ An instrument approach helps istics. A comparison of actual to facturers near an airport. a pilot land an aircraft in poor predicted runway length will Population density (PopDen), weather. A localizer or an identify those airports with taken from the 1980 U.S. Cen­ instrument landing system (ILS) shorter than expected runways, sus, is useful because an airport approach are two specific types given the characteristics of the in a less densely populated area of instrument approaches. communities they serve and the might be expected to have better Because of the inherent air traffic that uses the airport. facilities than one in a more inaccuracy of the landing THE REGRESSION ANALYSIS densely populated area, where systems and the vast amount of The regression analysis uses other airports would tend to be pilot navigational inputs, these runway length as the dependent closer. systems tend to be the most variable and combinations of the Taken together, these first 10 accurate and, therefore, the most other variables as independent variables describethe population desirable at a typical general variables. A regression analysis

Published by Scholarly16 Commons, 1992 JAAER, Winter 1992 3 Journal of Aviation/Aerospace Education & Research, Vol. 2, No. 2 [1992], Art. 6 Runway Extension Planning Model

Table 1 Variables Used in the Regression Analysis

Number Name Description

1. County County population 2. City City population 3. PopSrv Population served by airport 4. %pop Percent county population growth 1970 to 1980 5. MfgEmp Employment in manufacturing firms with more than 100 employees 6. CouEmp County employment 7. #Bus Number of business establishments in the county 8. Payroll County payroll 9. MfgPct Variable 5 divided by variable 6 10. PopDen Population density of the county 11. RwyLen Airport runway length 12. 100LL Aviation gasoline available at airport 13. Repairs Airframe or engine repairs done at the field 14. Agri Agricultural services available 15. Charter Aircraft charter available 16. Inst Flight instruction available 17. Rent Aircraft rental available 18. SEBase Number of single engine aircraft based at field 19. MEBase Number of multiengine aircraft based at field 20. JtBase Number of jet aircraft based at field 21. HBase Number of helicopters based at field 22. OpCar Number of annual air carrier operations 23. OpTaxi Number of annual air taxi operations 24. OpGA Number of general aviation local operations 25. Opltin Number of general aviation itinerant operations 26. OpMil Number of military operations 27. OpTot Total annual operations 28. Light Lighting on field 29. Twr Control tower on field 30. Appch Instrument approach 31. LOC Localizer or ILS approach 32. Nearby Nearby airport serves much of the same population area 33. SecArpt Another airport is the primary airport in the area

determines how the factors listed regression analysis identifies later. the comparison provides in Table 1 are associated with only airports that would be valuable information. If a runway different runway lengths at expected to have a runway of a is being considered for exten­ Alabama airports. Once this certain length. An element of sion. it would be useful to know. information is generated, the judgment, however, is missing in for example. that other airports residuals can be used to identify such an analytical comparison. with similar characteristics would airports that have shorter than Although the expected length of be expected to have longer run­ expected runways. the runway is not by itself a ways. The regression analysis Taking into account the recommendation for a longer supplies this type of information. variables from Table 1, the runway. as will be discussed Table 2 gives the results of https://commons.erau.edu/jaaer/vol2/iss2/6 DOI: https://doi.org/10.15394/jaaer.1992.1068JAAER, Winter 1992 4 Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi Runway Extension Planning Model

several regression equations. Table 2 shows that the tion. Fixing causation is Five equations use runway demographicvariable included in important when trying to draw length as the dependent vari­ the Selected Variables model is policy implications. able. The results of the tests of (%Pop) the percent population Note that the estimated significance (E-test) of each growth of the county in which coefficients tend to be more model indicate that the equa­ the airport is located. Although accurate in the Selected tions are found to be useful for %Pop was significant at the regression than in the All estimating runway length. The .E Q<.05Ievel in the All regression, regression because of the ratio in the first estimation is its significance fell as indicated probability of multicollinearity significant, but lower than the by a lower I-value. The %Pop among the variables. For others. This indicates a poorer fit variable's sign is negative, which example, an airport that supports due to the large number of insig­ could be interpreted as meaning a repair facility also tends to nificant variables. that counties experiencing lower have fuel available, and air The first, the full model, used population growth were assoc­ charter is often associated with a all other variables as indepen­ iated with longer runways firm that supplies rental and dent variables, and the coeffi­ (typically rural low population instructional services. cients are reported under the density regions). The multicollinearity is true column labeled All Variables. Three variables describing with based aircraft; if multiengine Running a regression analysis airport services were included in aircraft are based at a field, with all independent variables will the equation: availability of single-engine aircraft are likely to explain the greatest amount of charter services (Charter), flight be based there also. The variation in runway lengths. instruction (Inst), and the Selected regression includes Multicollinearity does exit in the existence of a control tower only the multiengine category. full model. However, if one is aw. An airport that provides This type of aircraft is most likely primarily interested in the predic­ charter services is associated to utilize a longer runway. The tion of runway length rather than with a runway 620 feet longer MEBase variable indicates that a interpretation of the individual than one without such a service; runway tends to be 55 feet effects of the independent the existence of flight instruction longer for each multiengine variables, multicollinearity should suggests a runway 760 feet aircraft based there, and the present no problem (Mendenhall, longer than one without such a variable is significant at better et al., 1986). Therefore, despite service; the existence of a than the Q<.01 level. the model's multicollinearity, the control tower, however, is The final variable included, result of the All Variables associated with a runway more SecArpt, indicates whether the regression model is a source of than 1,725 feet longer. airport is a secondary airport in useful information. The Inst variable is good for the area. If it is the runway would The next column, Selected illustrating that correlation does be expected to be 1,249 feet Variables, in Table 2 presents not mean that one variable shorter. the results of a stepwise necessarily causes another. For Dropping variables out of the regression procedure starting example, having a fixed base initial equation to estimate the with all 32 independent variables operation (FBO) on the field to Selected regression lowered the to simplify the full model. supply flight instruction might explanatory power somewhat. Eliminating variables that encourage the airport owner to The B Square in this equation is contain similar information lengthen the runway, so flight .627; in the initial All equation it (highly interrelated variables, instruction could affect runway was .813. While the Selected e.g., the multiple population length; but the causation could equation may be more desirable variables) and those exhibiting run the other way, too. FBOs in some ways, the additional low levels of significance might choose to locate at air­ explanatory power of the initial emphasizes the more important ports with longer runways. The equation suggests evaluating variables. Twenty-six variables causation could run either way, both equations to consider were dropped from the full model or another variable might be the runway length. to produce the simplified model. main factor causing the correla- Three other regression equa-

Published by Scholarly18 Commons, 1992 JAAER, Winter 1"992 5 Journal of Aviation/Aerospace Education & Research, Vol. 2, No. 2 [1992], Art. 6 Runway Extension Planning Model

Table 2 Regression Resufts: Independent Variable • Runway Length

Variables Included Independent All Selected Demographic Airport Operational

County 0.054 (2.88)*** City -0.019 (2.48)** PopSrv 0.030 0.018 (4.93)*** (7.61)*** %pop -21.354 -17.691 (2.21)** (2.09)** MfgEmp -0.019 (0.16) CouEmp -0.122 (2.49)** #Bus 0.920 (1.98)* Payroll -0.005 -1.052E-03 (2.71)*** (3.31)*** MfgPet 4.117 (0.25) PopDen -8.893 (1.70)* 100LL -219.020 (0.77) Repairs 449.46 (1.45) Agri 401.228 453.981 (1.83)* (1.94)* Charter 765.934 619.563 838.581 (2.35)** (1.91)* (2.59)** Inst 878.757 760.359 667.260 (1.73)* (2.94)*** (2.61)*** Rent -31.075 (0.06) SEBase -16.979 (1.56) MEBase 104.785 55.398 52.667 (2.91)*** (3.32)*** (3.18)*** JtBase -282.353 (2.02)** HBase -191.172 (1.89)* OpCar -0.241 (0.55) OpTaxi -0.014 (0.03)

(continued) https://commons.erau.edu/jaaer/vol2/iss2/6 DOI: https://doi.org/10.15394/jaaer.1992.1068JAAER, Winter 1992 "19 6 Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi

Runway Extension Planning Model

Table 2 , continued Regression Resuns: Independent Variable • Runway Length

Variables Included Independent All Selected Demographic Airport Operational

OpGA -0.217 (0.51) Opltin -0.238 0.088 (0.56) (8.10)*** OpMiI -0.223 0.020 (0.52) (2.50)** OpTot 0.216 (0.51) Ught 647.939 (1.08) Twr 1,769.683 1,724.822 1,884.254 (1.96)* (2.72)*** (2.98)*** Appch -259.459 (0.90) LOe -1,334.000 (1.68)* Nearby -442.546 -403.080 (2.00)** (1.82)* SecArpt -804.188 -1,248.963 -1,304.799 (1.62) (2.42)* (2.53)** Constant 2,744.780 3,601.062 3,758.181 3,445.848 3,586.393

BSquare 0.813 0.627 0.436 0.634 0.458 ERatio 9.930 27.775 39.764 24.246 43.546

* Significant at better than the .1 level. NOTE: ** Significant at better than the .05 level. The !-values are shown in parentheses. *** Significant at better than the .01 level. See Table 1 for a description of the variables.

tions were also estimated, using those particular characteristics. factors. Even if an a~rport is a stepwise procedure to look at For example, how long would the expected to have a longer run­ particular factors affecting run­ airport's runway be, if only the way based on demographic con­ way length. The column headed demographic characteristics of siderations, but not on the Demographic Variables includes the area or the number of operations at the field, a longer demographic factors only in operations at the field were runway might not be warranted. estimating runway length, the considered? The coefficients are Of course, the airport might be column headed Airport Variables not discussed in detail here, but under-utilized precisely because includes only airport character­ the reader can evaluate them in its runway is too short. Thus all istics, and the column headed Table 2. The regressions were factors should to be taken into Operational Variables includes included in order to compare the account; obviously, the set of only operations into and out of predictions based on all factors five regressions will supply more the field. These equations can with those based on demo­ information than anyone be useful if one wants to isolate graphic, airport, and operations regression.

Published by Scholarly20 Commons, 1992 JAAER. Winter 1992 7 Journal of Aviation/Aerospace Education & Research, Vol. 2, No. 2 [1992], Art. 6 Runway Extension Planning Model

ANALYSIS the number of operations. residuals beyond the threshold Uslna the Residuals for The residuals must be consid­ in all but one of the regressions. Predicted Runway Length ered within the context of other These airports are candidates for If actual runway length is information about the airport, but further examination for possible compared to the runway length some discussion is warranted to runway extensions. predicted in the regression see how they might be applied. Another 16 airports exceeded equations, airports with shorter Using a threshold of 500 feet on the threshold in the demographic than predided runways can be the residuals, three of the five regression, and 11 more identified. Subtracting the equations in Table 3 suggest a exceeded the -500 foot threshold predicted length from the actual longer runway at Wilson Field. in the selected variables length gives the residual in the The longer runway would be equation. Further examination of regression; these are reported in expected on the basis of all these airports revealed that Table 3. Negative residuals variables, demographics, and the some are secondary airports, so indicate a predicted runway number of operations at the that longer runways are readily length in excess of the actual airport. The Airport residual, available to aircraft requiring runway length. Therefore, a however, suggests the runway them. Although more detailed negative residual is an indication length is close to appropriate, analysis is needed to actually that a longer runway would be and the Selected residual recommend a runway extension, expected at the airport. suggests that the runway is too the statistics provide a guideline The residuals in Table 3 are long by more than 319 feet. The for identifying airports that have reported to the nearest foot. mixed evidence from the statis­ a shorter runway than would be Looking at airport number 6, for tical analysis does not support a expected, so warrant further example, Wilson Field would be runway extension for Wilson examination. Table 4 lists the 22 predicted to have a runway 753 Field. airports with residuals below feet longer than its actual length Table 3 reports 22 airports in -500 in all or all but one of the based on all of the variables, 319 which the residual is above the estimations. The numbers feet shorter based on the threshold (that is, the residual is associated with the airports are selected variables, 1,547 feet less than -500) in every the numbers from Table 3. The longer based on the demo­ regression equation or in every next section discusses a few of graphic variables, 92 feet longer equation but one. Eight have the airports named in Table 4 to based on the airport variables, residuals below -500 in all show the reader how the model and 1,548 feet longer based on equations, and another 14 have results can be used.

Table 3 Residuals From the Regression Analysis Variables Included

Airport Name All Selected Demographic Airport Operational

1. Lucky Field -95 -1,487 -1,932 -1,947 -2,080 2. Flomaton -1,202 -1,572 -1,859 -1,193 -1,754 3. Grove Hill Municipal -1,469 -1,611 -1,873 -1,118 -1,837 4. Shields 202 219 -1,585 262 -1,586 5. Roy E. Ray -645 -2,152 -196 -1,763 -1,674 6. Wilson Field -753 319 -1,547 -92 -1,548 7. Madison Sky Park 228 -2,082 -2,958 -1,530 -1,599 8. Huntsville-lacey's Sp. -791 -1,941 -3,200 -1,583 -1,757 9. Red Bay Municipal -272 -974 -1,506 -743 -1,313 10. Ware Island -550 -819 -1,854 -643 -1,222 11. North Mobile County 348 -1,046 263 -1,151 -1,212 (continued) https://commons.erau.edu/jaaer/vol2/iss2/6 DOI: https://doi.org/10.15394/jaaer.1992.1068JAAER, Winter 1992 21 8 Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi

Published by Scholarly Commons, 1992 9 Journal of Aviation/Aerospace Education & Research, Vol. 2, No. 2 [1992], Art. 6 Runway Extension Planning Model

Table 3 - continued Residuals From the Regression Analysis Variables Included

Airport Name All Selected Demographic Airport Operational

56. AUbum-Opelika -212 -1.102 -833 -1.478 -825 57. Geneva Municipal -865 -794 -58 -549 286 58. St. Elmo 216 664 1.524 503 370 59. Scottsboro Municipal 431 830 -301 443 276 60. Butler-Choctaw County -233 -259 118 -28 323 61. Weedon Field -180 -254 -36 -171 5 62. Stevenson-Bridgeport -20 314 302 12 186 63. Robbins Field -37 1.102 -184 629 418 64. Blackwell Field 451 -259 180 430 -505 65. Bibb County 473 843 189 754 429 66. St. Clair County 102 289 -49 -282 330 67. Camden Municipal 142 431 302 754 578 68. Isbell Field -711 -552 -302 -998 508 69. Thomas C. Russell Field -84 29 391 441 446 70. Coosa County 1.677 968 636 1.407 864 71. Pine Hill Municipal 264 731 591 1.054 896 72. Frank Sikes 757 1.067 615 1,051 439 73. Lee Merkle Field -834 -472 496 -653 256 74. Clayton Municipal -315 446 1,001 75 1,159 75. Albertville Municipal 337 -419 806 -328 828 76. Walker County -376 497 85 371 331 77. Richard Arthur Field 599 711 738 684 1.132 78. Atmore Municipal 518 742 938 761 808 79. Jackson Municipal 1.297 1.374 906 1.870 1.186 80. George Downer 967 1,414 1,078 1,018 1.208 81. Troy Municipal -4 -1,772 895 -2,101 32 82. Andalusia-Opp -733 52 700 -108 1.157 83. Moton Field 231 780 927 887 1,150 84. Demopolis Municipal 898 1,491 964 1.100 1.194 85. Posey Field 352 869 1,069 571 1.194 86. Pryor Field -431 307 -81 353 -967 87. Brewton Municipal 577 242 1,113 499 345 88. North Pickens 1,365 1,650 1,280 1,704 1,528 89. . Fairhope Municipal 596 680 1,377 546 716 90. Folsom Field 643 982 425 878 822 91. Bessemer -45 -95 355 309 271 92. Monroe County 545 1.512 1,860 1.573 2.224 93. Talladega Municipal 566 867 1,816 631 1.090 94. Tuscaloosa Municipal -527 337 778 63 1.060 95. Muscle Shoals 37 1,054 443 497 100 96. Gadsden Municipal 394 1.733 1,699 1,608 2,878 97. Marion County 1,408 2.938 2,644 2.275 3.150 98. Anniston-Calhoun County 548 1,593 1,726 1.775 2,263 99. Huntsville-Madison Co. 349 -380 1,893 -118 1,151 (continued) https://commons.erau.edu/jaaer/vol2/iss2/6 DOI: https://doi.org/10.15394/jaaer.1992.1068JAAER, Winter 1992 ·23 10 Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi

Further Analysis Table 4 runway of 3,150 feet exceeds the A closer examination of the Airports with Short Runways -500 foot threshold in two of the airports listed in Table 4 reveals According to Regression regressions, combining the that 12 of the 22 listed are Results statistical information on the two privately owned airports. The Airport Name airports with other factors privately owned airports are suggests that a runway exten­ Lucky, Flomaton, Ray, Madison 1. Lucky Field sion would be more feasible at Sky Park, Huntsville-Lacey, Ware 2. Flomaton the Lanett airport than at the Island, Sky Harbor, Ardmore, 3. Grove Hill Municipal Valley airport. The point is that, Valley, Jacksonville, Martin, and 5. Roy E. Ray to arrive at policy recommenda­ Huntsville North. Extending the 7. Madison Sky Park tions. other factors must be runway at any of these or at 8. Hunstville-Lacey's Spring considered along with this publicly owned airports in that 9. Red Bay Municipal statistical information. The area might be feasible unless a 10. Ware Island analysis suggests where action nearby airport has a longer 12. Sky Harbor might be desirable. runway. Any recommendation 14. Hazel Green For some airports, the informa­ would require further analysis. 15. Ardmore tion from the statistical analysis For example, in Lanett the Valley 17. Valley can be used to confirm the desir­ Airport's runway is 2,950 feet, at 20. Jacksonville Municipal ability of extending a runway. For least 772 feet shorter (and 21. Headland Municipal example, the management of the sometimes more than 1,893 feet 22. Wetumpka Municipal Auburn-Opelika airport (Table 3, shorter) than predicted in the 25. Chatom Municipal airport number 56) would like to regression equations. These 26. Brundige Municipal extend a 4,OOO-foot runway to statistics alone suggest a runway 27. Martin Field accommodate larger aircraft. The extension. But circumstances not 31. Guntersville Municipal statistical analysis shows that reflected in the statistics must be 35. Enterprise Municipal when the airport, operational, considered. First, that the field is 50. Huntsville North and demographic characteristics privately owned is not an 56. Auburn-Opelika are taken into account. the argument against a runway runway is shorter than would be extension, but it complicates the expected. One would not want to State's role. Second, because extend a runway solely because the runway is in a bend in the Municipal Airport is nearby and of this statistical analysis, but it is Chattahoochee River an exten­ could be extended more easily. comforting to find out that the sion would be almost impossible. Although Lanett Municipal's statistical analysis confirms that Third. the publicly owned Lanett (Table 3, airport number 28) this airport which wants a longer

Published by Scholarly24 Commons, 1992 JAAER, Winter 1992 11 Journal of Aviation/Aerospace Education & Research, Vol. 2, No. 2 [1992], Art. 6 Runway Extension Planning Model

runway appears to be a candi­ expected. Only one airport. CONCLUSION date when considering its demo­ Brewton (number 87 in Table 3) This paper has reported the graphic and airport character­ was identified as a strong method and results of a study istics relative to other airports candidate for a repair facility in evaluating the length of runways around the state. this regard. Although the State at airports in Alabama. Using Ukewise. data that suggests a may not be interested in taking runway length as the dependent runway is relatively long an active role here, the analysis variable. a collection of inde­ compared to others in the state did identify a possibility for local pendent variables describing the need not preclude an extension. development. airport characteristics and If an airport operator wants to An instrument approach aids demographic characteristics of extend a runway, state officials aircraft landing during poor the area surrounding the airport should consider the data but be weather. conditions and at night. was used to predict runway able to identify special circum­ The existence of an instrument length. The residuals from the stances that might make an even approach was also used as the regressions were examined in longer runway desirable. In dependent variable in a order to identify airports with short. the statistical analysis is regression analysis to identify runways that were shorter than the basis for recommendations: five airports that would have expected. The regression it can only show how a particular been expected to have instru­ analysis identified 22 airports airport compares to others ment approaches. Many states that could be candidates for around the state. Officials can fund. own. and operate. longer runways. Of those 22. a use the comparison along with instrument approaches as well further analysis. only partly other information to decide as other navigational aids described in the paper. led to a whether to extend a runway. (NASAO. 1988). Information from recommendation that 9 airports In the complete technical the analysis could aid in the be considered for runway report from which this paper is identification of candidate extensions. developed, the regression airports for possible instrument Regression analysis was used analysis and other information approach installations. to evaluate other airport charac­ about Alabama airports were the In another regression. based teristics as well, such as the basis of a recommendation to aircraft was used as the existence of a repair facility at consider runway extension at dependent variable in similar the airport and an instrument nine airports. The Auburn­ analysis. Three airports were approach. The method used Opelika airport described above identified as having fewer aircraft here can be applied to other is representative of an airport for than expected by two standard types of transportation to identify which the regression analysis deviations, and two had more facilities that are candidates for confirms that a runway extension than expected. Airports with development. taking into account would be desirable. fewer aircraft than expected a large number of factors at Other Applications could be candidates for develop­ similar facilities. The preceding sections have ment. Those with more than Aviation capital investment illustrated how a regression expected could perhaps be used decisions are theoretically model can be used to identify as models for expanding under­ justified by some form of a airports with runways shorter developed airports. benefit-cost analysis to support than would be expected. In the The previous sections of the the project's feasibility. The larger report being made for the paper described in detail how Federal Aviation Administration State. the same type of model runway length can be analyzed published a model guide in 1982 was used to look at other factors with a regression model. (FAA. 1982) to provide a sys­ as well. The availability of repairs Undoubtedly this type of analysis tematic approach to answering at the airport was used as the could be adapted to other economic questions for federal dependent variable in similar transportation issues as well. aviation projects. An evaluation analysis in order to identify such as channel depth of water­ (McLeod. 1984) found that the airports without repair facilities ways and number of lanes of guide is not generally used. The where they would have been traffic in an urban beltway. authors' experience also https://commons.erau.edu/jaaer/vol2/iss2/6 DOI: https://doi.org/10.15394/jaaer.1992.1068JAAER, Winter 1992 25 12 Holcombe and Burdg: A Method for Identifying General Aviation Airports that are Candi Runway Extension Planning Model

suggests that a large proportion Allocation of funds is often taking into account the air traffic of state aviation development based on political pressures and community characteristics projects are undertaken without rather than on analytical findings. they serve. However, if state or the support of a benefit-cost The analysis undertaken here federal funds are available for evaluation. can help identify airports that runway extensions in a state, this One reason benefit-cost could benefit most from a type of analysis is very appro­ analysis may not be universally runway extension. priate because it can help used is that a state's decision to Because the methodology identify those airports with the spend money can depend more simply compares airports and greatest relative need. It is not a upon the availability of funds identifies those that have substitute for benefit-cost than upon an assessment of relatively short runways consider­ analysis, but rather a comple­ need. If no money is available for ing the.populations they serve, ment that can help identify those runway extensions, then none no conclusions can be drawn airports at which the net benefits will be undertaken regardless of about cost-effectiveness of from runway extension would be the desirability of extensions; extending a runway. This analy­ the greatest. conversely, if money is available, sis simply identifies airports with government will tend to spend it. relatively short runways when

Randall G. Holcombe received his Ph.D. in economics from Virginia Polytechnic Institute and taught economics at Texas A&M University and at Auburn University before taking his current position of Professor of Economics at Florida State University, Tallahassee, Florida. He is author of four books and numerous articles on economics and public policy. Henry B. Burdg earned his Bachelor of Aviation Management and M.B.A. degrees from Auburn University and currently is Director of the Auburn Technical Assistance Center (ATAC). A Certified Management Consultant (CMC) he has worked as an airport planner and has authored several published articles. REFERENCES

Ashford, N., & Wright, P. H. (1979). Airpon Engineering. New York: John Wiley & Sons. Federal Aviation Administration. (1972). Planning the State Airport System (AC 150/5050-3A). Prepared by the Federal Aviation Administration in association with the National Association of State Aviation Officials, Washington, DC: U.S. Government Printing Office. Federal Aviation Administration. (1975). Advisory Circular-Utility Airpons, Air Access to National Transportation (AC 150/5300-4B Changes 1-8). Washington, DC: Author. Federal Aviation Administration. (1982). Economic Analysis of Investment and Regulatory Decisions-A Guide (FAA-APO-82-1). Washington, DC: Author. Federal Aviation Administration. (1983). Advisory Circular-Airport Design Standards-Transport Airpons (AC 150/5300-12). Washington, DC: Author. Horonjeff. R., & McKelvey, F. X. (1983). Planning and Design ofAirports (3rd. ed.). New York: McGraw-Hili. McLeod, D. S. (1984, January). Evaluation of FAA's Economic Analysis Guide. Paper presented at the annual meeting of the Transportation Research Board, Washington, DC. Mendenhall, W., Reinmuth, J. E., Beaver, R., & Duhan, D. (1986). Statistics for Management and Economics (5th. ad.). Boston: DuxbUry Press. National Association ofState Aviation Officials. (1988). The States andAir Transportation. Silver Spring, MD: NASAO Center for Aviation Research and Education. •••

Published by Scholarly26 Commons, 1992 JAAER, Winter 1992 13