Evaluating Methods for Counting Aircraft Operations at Non-Towered Airports (2015)

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.

6available literature resulting from national and international aviation conferences, all of which was scrutinized for salient points, common elements, and best practices. The literature review revealed two recent efforts at record- ing non-towered airport operations. One was entitled X-band Radar: Radar for Monitoring the Airport Environs (Hartfiel 2011). This report was a student project for submission to the FAA Real World Design Competition. It investigated the uti- lization of X-band off-the-shelf commercial radar typically used for a private watercraft, along with associated hardware and software, to collect aircraft movement data. Unfortunately, the project team did not receive FAA and Federal Communi- cation Commission (FCC) approval for the installation of the X-band radar equipment so it was not identified as a current technology to evaluate. One special concern revealed in the study was the power and spectrum output of the radar and the potential for interference with aircraft and air traffic control operations. Another effort, also a result of FAA Real World Design Competition, was titled Aviation Operations Monitoring System (Angelini 2011). It proposed the installation of Radio Frequency Identification (RFID) transmitters on general aviation aircraft along with airport installation of RFID readers in order to record aircraft operations. (Note: General aviation includes all segments of flying except for airlines and military.) The biggest issue for this technology is that it required the installation of RFID transmitters in all the aircraft, which presents similar challenges as found with ADS-B in that most general aviation aircraft owners would not voluntarily install the equipment. This is especially true in this case where no benefit to the owner would result from the installation. While this system reportedly would accurately count aircraft passing within its range of view, it would have similar challenges to other systems in deter- mining the phase of operation and actual counting appli- cability. For these reasons, it also was not included as a current technology to evaluate. The research project included five primary tasks: 1. Developing an amplified work program and identifying subsequent advances in aircraft operations counting tech- nologies since the publication of ACRP Synthesis 4: Counting Aircraft Operations at Non-Towered Airports. 2. Contacting states, airports, and MPOs about their count- ing practices. 3. Developing the test program. 4. Conducting the test program. 5. Producing a final report. These tasks are outlined in greater detail in the following pages and sections of this report. Tasks Task 1: Developing amplified work program and identifying subsequent advances in aircraft operations counting technologies since the publication of ACRP Synthesis 4. Task 1 involved reviewing the original work program pro- vided during the request for proposals and updating it based on committee comments as necessary. After the amplified work program was accepted, ACRP Synthesis 4 was reviewed to glean information about previous research performed by TRB and others on the research topic. Then, a literature review was conducted to identify any new technologies and methods that had been implemented subsequent to the publication of ACRP Synthesis 4. It was not the intent of this study to redo this literature review, but rather to add to it the review of any literature that had occurred since ACRP Synthesis 4, or that may have become available through more powerful research engines. The updated literature review included a search of the academic and research literature related to counting air- craft traffic operations at non-towered airports, as well as the C H A P T E R 2 Research Approach

7 The most widespread use of the operations data was for inclusion on the airport’s FAA Airport Master Record Form 5010. Aviation forecasting and justification for airport devel- opment projects were also listed as some of the most com- mon uses of the data. The contacting initiative revealed that, generally, states, planning entities, and airports are interested in a user-friendly, accurate, and cost-effective counting/estimating methodol- ogy. While sampling operations with an acoustical counter ranked as the favored method of the FAA (FAA 2007), only five entities utilized this method. In fact, many state aviation officials had abandoned using acoustical equipment because either they felt they weren’t accurate, they couldn’t verify their accuracy, the equipment was old, or it was in a state of disrepair. The contacting initiative also revealed that most entities that take sample counts do not use a statistical process for extra- polating the results into an annual estimate. Rather, they use a seasonal or monthly adjustment factor based on towered air- port operations data. Several entities estimate operations by use of an estimated OPBA method, but there is no consistent base value used, as they ranged from 100 to 750 operations per based aircraft. While VID systems were thought to offer the most informa- tion about the aircraft traffic, including aircraft identification numbers and aircraft make and model (which is most useful when critical aircraft need to be identified for justification of runway extensions), none of the responders in the contacting initiative indicated use of this system for counting purposes. Airports that had installed VID systems did so for other rea- sons. One consideration as to why VID systems are not being used to count aircraft traffic is the cost. While they provide information on aircraft identification numbers, make, model, weight, operations, etc., they are costly (anywhere from $50,000 to over $150,000 depending on airport layout) and require a monthly charge from the service provider to obtain the data. Task 3: Developing the Test Program The test program developed in the amplified work program was updated based on the results of the contacting initiative. The completed literature review and contacting initiative did not uncover any significant new information about airport operation counting and estimating methods being practiced today except for the use of the IFPTO method, which was added to the research. Based on the requests of the ACRP Project 03-27 panel, findings from ACRP Synthesis 4, the literature review, the contacting initiative, and the budget allocated for the proj- ect, three methods of estimating annual operations and four counting technologies were advanced for testing. The meth- ods to estimate annual operations included (1) multiplying Task 2: Contacting States, Airports, MPOs After the literature review was performed, the aviation offi- cials for each state within the United States were contacted (contacting initiative) by phone and asked to fill out an online survey regarding their practices for counting traffic at non- towered airports. They were also asked if they knew of any entities within their state that were conducting their own air- port traffic counting program and, if they did, those entities were also contacted. In some cases the individuals contacted did not want to participate in the online survey, so if will- ing, their information was collected during a phone call or by information on their respective entity’s website or other pub- lished information (e.g., state aviation system plan). If their responses revealed a new method of counting aircraft from those identified in ACRP Synthesis 4, they were subsequently contacted directly for more information. The contacting initiative revealed that, by and large, the vast majority of state aviation officials collect traffic counts by simply asking the airport manager, fixed-based operation (FBO), or other entity associated with the airport what the annual operations are, which typically consists of an edu- cated guess. Many entities base their estimates on a valid data source, but the data source itself does not contain all aircraft so it intrinsically excludes traffic (e.g., Flight Aware or pilot sign-in log books). A new method for estimating aircraft operations was revealed in the contacting initiative: one airport estimated its total operations by adding three VFR flights to each instru- ment flight rules (IFR) flight plan for the year to produce an annual estimate, but admits the ratio is an estimate. This method is referred to in this report as the IFPTO method from this point forward. The contacting initiative also revealed how states and other entities were using operations data, which included a variety of reasons: • Airport control tower justification • Airport development project justification • Airport Master Record (Form 5010) • Airport operations fleet mix • Aviation forecasts • Budget funding justification • Community relations • Activity changes (increasing/decreasing) at airports • Economic impact statements • Environmental Assessment or Impact Documentation • Instrument approach procedure development justification • Lease or curfew compliance • Master plan updates • Measure of performance • Runway pavement life spans

8based aircraft by an estimated number of OPBA, (2) applying a ratio of FAA IFPTO, and (3) expanding a sample count into an annual estimate through extrapolation. In order to expand a sample into an annual count through extrapola- tion, the sample first has to be taken. This can be done by a person physically counting aircraft operations, but is more typically done by some type of aircraft counting technol- ogy. The different aircraft counting technologies (to obtain the sample count) that were advanced to the evaluation stage to determine their ability to count aircraft traffic at non-towered airports included acoustic (both automated acoustical and sound-level meter acoustical), security/trail cameras, and VID with a transponder receiver. These meth- ods and technologies are described in detail in the following sections. The airports where the estimating methods were tested included multiple non-hub airports with FAA VFR towers with less than approximately 730 air carrier operations per year. (See Chapter 3 for information of these airports.) The airports where the counting technologies (i.e., equipment) were tested included Purdue University Airport (LAF), Indi- anapolis Executive Airport (TYQ), Paoli Municipal Airport (I42), and Eagle Creek Airpark (EYE). A summary of the methods and locations is shown in Table 2-1. Estimating Methods The three methods of estimating annual operations advanced for testing included OPBA, applying a ratio of the IFPTO, and extrapolating from a sample count. These methods were tested by climatic region using a dataset derived from FAA towered airport records. The dataset is described here: Dataset Sources: The data sources for this analysis were the FAA Terminal Area Forecast (TAF) and the FAA Operations Network (OPSNET) databases from 2006 to 2010. The TAF includes his- torical and forecast statistics on passenger demand and aviation activity at U.S. airports. The TAF contains historical and forecast data for enplanements, airport operations, and based aircraft. Once published the TAF remains constant until its next publica- tion with the only exceptions being significant traffic shifts by major airlines, or the revelation of a significant historical data error. This database was used for enplanements. To augment the TAF, OPSNET data was used for operations because OPSNET is the official source of National Airspace System air traffic operations and delay data. OPSNET does not include data on enplanements, which is why the TAF was also used. Note that towered airports actively count and report opera- tions data via the air traffic controllers, while non-towered typically do not. To test these three methods described above, estimated operations must be compared to actual operations. This can only be done where annual operations are known, which are at towered airports. Therefore, TAF and OPSNET data for certain small, towered airports was used in this part of the research as a proxy for non-towered airports. Small, towered airports were defined, for this study, by the following criteria: • Non-hub public use airport with FAA VFR tower or FAA con- tract tower; • Less than 10,000 annual enplanements (i.e., non-primary air- ports); and • Less than 730 air carrier operations per year (i.e., an average of one air carrier flight per day). Table 2-1. Counting methods tested and locations summary. METHOD A. Equipment - Sample Count: Expand sample count to annual count or count full year. B. Operations Per Based Aircraft (OPBA) Estimate annual operations as a product of based aircraft. C. Instrument Flight Plans to Total Operations (IFPTO) Estimate annual operations as a ratio of flight plans filed. 1A Sound-Level Meter Acoustical Counter. Test Locations: LAF, TYQ, EYE, I42 FAA as source for historic tower counts and based aircraft. FAA as source for historic tower counts and instrument flight data. 1B Automated Acoustical Counter Test Locations: LAF, TYQ, EYE, I42. Test Locations: Multiple FAA VFR Towered airports across the U.S. Test Locations: Multiple FAA VFR Towered airports across the U.S. 2 Security/Trail Camera Test Location: TYQ, EYE, I42. 3 Video Image Detection - Test Location: TYQ. 4 Transponder Receiver - Test Location: TYQ. Prepared by: Woolpert, Inc.

9 objective of analyzing this method was to determine if there was a number that could be used by airport management/ planners to multiply by their number of based aircraft to obtain a relatively accurate estimate of annual airport operations. To be evaluated, the operations estimated using the OPBA method had to be compared to airports with known annual traffic and based aircraft. As stated previously, since valid operations data do not exist for non-towered air- ports, small, towered airports (i.e., STAD) were used as a proxy for the comparison. Each airport’s recorded opera- tions for five historic years were divided by the number of based aircraft at the facility for the same five historic years (2006–2010). The states were divided into their nine climate regions defined by the NOAA, National Climatic Data Cen- ter and the airport’s region noted. The population of the associated city of the airport was also determined by access- ing U.S. Census data and noted. The research team then determined if there was a consistent number(s) of OPBA that occurred at these facilities and if it varied by climate or population. Non-hub airports were chosen because they are more likely to not have a large amount of commercial service and would better reflect non-towered airports. Within this grouping, FAA VFR tower and contract tower airports were chosen because these airports were assumed to more closely resemble non-towered airports than towered airports that were busy enough to war- rant the installation of radar. Airports with less than an average of 730 air carrier operations were chosen because 730 would represent approximately two air carrier operations per day (one takeoff and one landing) and this would more closely represent non-towered airports than airports that experience regular and consistent air carrier operations. The application of this criteria resulted in 205 airports being included in what this research labeled the STAD. Climate data was obtained from the National Oceanic and Atmospheric Administration (NOAA), National Climatic Data Center. Through their climate analysis, they have identified nine climatically consistent regions within the contiguous United States, which were used in this analysis. (See Table 2-2.) OPBA. When using this method, operations are pro- jected as a product of based aircraft by multiplying an air- port’s based aircraft by an estimated number of OPBA. The State Climatic Region State Climatic Region AK Alaska MT West North Central AL Southeast NC Southeast AR South ND West North Central AZ Southwest NE West North Central CA West NH Northeast CO Southwest NJ Northeast CT Northeast NM Southwest DE Northeast NV West FL Southeast NY Northeast GA Southeast OH Central HI Hawaii OK South IA East North Central OR Northwest ID Northwest PA Northeast IL Central RI Northeast IN Central SC Southeast KS South SD West North Central KY Central TN Central LA South TX South MA Northeast UT Southwest MD Northeast VA Southeast ME Northeast VT Northeast MI East North Central WA Northwest MN East North Central WI East North Central MO Central WV Central MS South WY West North Central Prepared by: Woolpert, Inc. Table 2-2. States and NOAA climatic regions.

10 tion, some entities use a seasonal or monthly adjustment fac- tor to expand an airport’s sample into an annual estimate. For example, an entity believes that fifteen percent (15%) of the year’s operations happen in July because that is the average for all the towered, general aviation (GA) airports in its state. If it samples traffic at a non-towered airport for the month of July, its sample would account for 15% of the total annual operations and it would compute its total estimated opera- tions accordingly. To test the monthly/seasonal adjustment factor, the per- centage of operations that occurred in each month was cal- culated for each region based on the airports selected in the previous exercise. The random samples of operations for the same sample periods identified in the statistical extrapolation exercise described above were then extrapolated using these monthly percentages. The research team then compared the estimated operations to the actual operations. Counting Technologies As stated earlier, in order to expand a sample into an annual count through extrapolation (be it through statistical or monthly/seasonal adjustment factor extrapolation), the sam- ple first has to be taken. This is often done through the use of some type of technology designed to count aircraft operations. This research looked at four different types of technologies: acoustic, security/trail cameras, VID, and ADS-B transponder receivers. The acoustic aircraft counters are designed to record takeoffs, but not landings. The underlying assumption is that for every takeoff there is a landing, so a total count is produced by doubling the takeoffs recorded. The security/trail cameras record traffic that passes in front of them and the images are then manually tallied. The VID equipment works on the same principle, but is automated and requires an annual service contract. It captures an image of the aircraft N-number as it passes by the camera and the service provider analyzes the image and provides detailed information about the aircraft. To augment the capability of its VID system, the service provider for the equipment tested in this study included an option for a simple transponder receiver programmed to detect ADS-B and Mode S (transponders that support Traffic Collision Avoid- ance System) transmissions that met certain criteria. This identified equipment was evaluated in a multiple case study using four airports. A long-term study of all the equipment was performed at TYQ, where the equipment was left in place for approximately 7 months to determine its durability. The study also determined how the equip- ment would perform in both warm and cold months and how much information could be stored before data had to be downloaded. Short-term accuracy tests were also performed at TYQ, LAF, I42, and EYE for the acoustic counters and TYQ, I42, and EYE for the security/trail cameras. Because of the expense of IFPTO. When using this method, operations are pro- jected as a ratio of IFPTO. Since IFR operations are tracked and recorded by the FAA for all airports (VFR operations are not) and total operations are known for towered airports, total operations could theoretically be estimated from IFR opera- tions if a consistent ratio existed between them. The objective of analyzing this method was to determine if there was a ratio of IFR to VFR operations that could be used by airports to obtain a relatively accurate estimate of annual airport operations. To test this method, annual IFR traffic for the same small, tow- ered airports (i.e., STAD) described in the OPBA exercise were compared to their total traffic using the same dataset sources described above. Each airport’s annual operations were divided by the number of IFR operations for the facility for five historic years (2006–2010). The climate region of the airport was also acquired and noted. The research team then determined if there was a consistent ratio of IFR flight plans filed to total operations that occurred at these facilities and if it varied by climate. Extrapolation. When using this method, operations are projected by expanding a sample count into an annual esti- mate. When sample counts of aircraft operations are taken at an airport, the counts are typically extrapolated into annual operations estimates using one of two different types of extra p- olation methods: (1) statistical extrapolation and (2) monthly/ seasonal adjustment factor extrapolation. Statistical Extrapolation. At many airports, activity will vary due to day of week, weather, and season. The goal is to take a sample(s) that captures these differences. Previous research (FAA-APO-85-7, Statistical Sampling of Aircraft Operations at Non-Towered Airports) has indicated that the most accurate and cost-effective way to do this is to sample traffic for two weeks for each of the airport’s seasons and extrapolate that into an annual estimate. However, not all entities do this. To test different sampling periods, random samples of daily historic 2010 tower operations from FAA OPSNET for different time- frames was gathered for one of the STAD airports in each of the nine climate regions and statistically extrapolated into annual estimates using the method in FAA-APO-85-7. The random samples included the following timeframes: 1. One week in each season (number of seasons depends on climate). 2. Two weeks in each season (number of seasons depends on climate). 3. One month in spring, summer, or fall. 4. One month in winter. The research team then compared the estimated opera- tions to the actual operations. Monthly/Seasonal Adjustment Factors. The contacting initiate revealed that rather than using statistical extrapola-

11 Limitations and Assumptions of the Study A comparison of estimated results to observed data is rou- tinely used as a way to measure accuracy of a model. In the case of annual airport operations, the only way to acquire observed data at non-towered airports would be to physically watch and record each operation for a full year. This practice is not feasible or practical, which is the underlying need for this current research project. Because of this limitation, the methods described herein for this current research project are often compared to small, VFR towered airport data. (For a description of the airports used in the analysis, please refer to dataset sources under Task 3.) (Note: VFR towers are air- port traffic control towers that provide takeoff and landing services only. They do not provide approach control services. Aircraft on IFR flight plans can still take off and land at an airport with a VFR tower.) When using small, towered airports in the analysis, an assumption is made that traffic behavior is similar from VFR towered airports to non-towered airports. In the past, towered airports were believed to have more consistent traffic due to more and better instrument approach procedures serving them. Any potential differences that may exist between towered and non-towered airports have theoretically narrowed over time with the introduction of global positioning system (GPS) guided instrument approaches to a vast spectrum of non- towered airports. Where the better instrument approaches may have normally been found at towered airports, GPS technol- ogy has increased the utility of most airports during inclem- ent weather, so weather conditions are less of a differentiator now as compared to years past, making the majority of publicly owned airports similar in approach capabilities. According to the FAA Global Navigation Satellite System (GNSS) Program Office, there were 2,664 Wide Area Augmen- tation System (WAAS) capable airports in the United States as of May 31, 2012, and 80% of these were non-Part 139 FAA certificated airports. Most GA, non-towered airports are also typically non-Part 139. (Note: WAAS provides augmentation information to GPS receivers to enhance the accuracy and reliability of position estimates and allows for very accurate instrument flight procedures into airports). While some of the 80% will have towers, the instrument approach differentiator between towered and non-towered airports has diminished significantly. It is also important to note that some of the data on the small, non-towered airport comes from the FAA TAF and OPSNET databases. The most recent year where the TAF included actual and not forecast data at the time this analysis was initiated was 2010. Therefore, all use of FAA TAF data included 2010 and earlier. To remain consistent, data used from FAA OPSNET also included only 2010 and earlier. leasing and installing the video imaging detection equipment and transponder equipment, it was only tested at TYQ. (Note: The FAA would not approve testing the VID or the security cameras at LAF because they would need to be located within FAA restricted set-backs to work.) The information from the installation process, durability study, and short-term studies were used to rate each piece of equipment on the following criteria: • Principle(s) of operation and intended use • Computer requirements • Data provided • Ease of portability • Durability • Ease of installation and airport impacts • Maintenance and operation • Ease of data retrieval • Performance in various weather and lighting conditions • Service contract requirements • Cost • Accuracy The data from the accuracy tests were compared to visual observations, which included recording each airport opera- tion, the aircraft N-number, aircraft type (e.g., single engine, multi-engine, turbine/jet, helicopter), the date, the time, and type of operation (i.e., landing, takeoff, touch-and-go.) These data were then compared to the data from each aircraft counter (or group of counters as appropriate) and differences noted. The percentage of error for each counter (or group of counters, as appropriate) was computed. It is important to note that the FAA determined that any equipment installation on the airport, even if it were tem- porary and outside the runway safety area, required an FAA approval (through the filing of an FAA Form 7460) in order to be in compliance with Title 14 of the Code of Federal Regula- tions (CFR) Part 77. In the case of this research, a Form 7460 was needed for each piece of equipment at each airport. The locations of the equipment are detailed in the results section and Appendix D, which include airport diagrams. Task 4: Conducting the Test Program This task involved implementing the research program developed in Task 3. The goals of the program were to evalu- ate the estimating methods and counting technologies at different airports, giving consideration to their unique char- acteristics where appropriate, and compare actual operations to estimated or counted operations. Task 5: Producing the Final Report This task involved the creation of this final project report.