Computer simulations are used in airspace and airport modeling in order to evaluate capacity and operations. Existing operations and future demand are used as inputs to these models in support of the decision-making process. The primary issue in the modeling process is safety. Safety cannot be compromised. Airport and airspace efficiency is secondary.
The NJIT Study Team assessed the various studies to date and reviewed the software models currently used by all stakeholders and endorsed by the FAA. In addition, the team reviewed models which are presently being tested and evaluated on behalf of or by the FAA for potential future. A description of each model follows.
SIMMOD, the FAA’s current airport and airspace simulation model, depicts the dynamic interaction of aircraft movements and is time-tested. This powerful software package, programmed in the language SIMSCRIPT II.5, simultaneously studies and evaluates en-route traffic, terminal area traffic, and ground operations at one or more airports. SIMMOD produces measures of airport capacity, aircraft travel time, aircraft delay, and aircraft fuel consumption. After a standard scenario has been established with data from existing or proposed operations, the input data may be changed to develop and evaluate different scenarios. This allows the "what-if" questions to be analyzed. The model may be implemented for projects as large as a major route network and as limited as individual terminal gate operations.
The SIMMOD analysis aids planning in the following areas:
Reports and graphs provide statistics on airport capacity, delay, fuel consumption, time and resource requirements for air and ground operations. Different airspace and airport scenarios may be compared based on this output.
SIMMOD, a discrete event model, represents a system that evolves over time but that changes at single, or discrete, points in time. Events (e.g., runway status, number of aircraft in a queue) occur at these points in time. These events are instantaneous in the model and make changes in the stated variables. SIMMOD events are processed in sequence based on the event schedule and the simulation clock. The model simply steps from one event to the next in the order of its appearance in the event schedule. If two events occur at the same time, SIMMOD will process the one with the higher priority first, without moving the clock forward. The priority is tracked via the Simulation Log. After each event is processed, SIMMOD checks for the next event. An event can cause other events to be added to the event schedule.
External events are defined directly by data input. These external events can generate a series of internal events. The three types of external events are:
The simulation clock is updated as events are processed. The clock starts at 00:00 and schedules the times of external events. SIMMOD advances the clock to the first event, updates the system based on the event, and adds any internal events generated by the event. The clock is advanced to the time of the next event and the process continues until it reaches the end of the list or a user-specified event that ends the simulation.
An airport and airspace system is represented by SIMMOD as a collection of nodes and links. A node is defined as a point in a coordinate system where SIMMOD evaluates an aircraft’s position with respect to other aircraft in the system. A link defines the path between two nodes. Therefore, the aircraft move from one node to another via a link. Ground and airspace nodes are separate groups. Ground nodes are gates, departure queues, runway or taxipath intersections. Airspace nodes are navigational fixes, holding queues, merge points, or interfaces with an airport. Ground and airspace links are also separate links. Ground links are runways or taxipaths. Airspace links are routes.
Random variables are used in SIMMOD to produce output that represents variations in air traffic. This stochastic modeling allows SIMMOD to create realistic results. This output gives a unique representation of daily variations in air traffic. Statistically significant tendencies may be established for a particular data set by running several iterations. Reports provide aggregate values, averages, and standard deviations where applicable. Random linear variables can be defined for certain phenomena to enhance the simulation. This is accomplished by user-defined cumulative probability distributions that determine the amount of variation. Distribution values may be employed for:
A sequence, or stream, of random numbers is used by SIMMOD for each iteration. The first number in the stream, the seed, defines the outcome of the ensuing random numbers. Therefore, using the same seed, the same simulation results can be recreated.
A "flight" is an aircraft with a unique identifier and a data set that includes: type of flight, starting time, and airspace route. The three types of flights are: arrivals, departures, and overflights. The arrival flight, which starts in airspace, consists of a flight that:
The arrival can turn around and become a departure flight. Therefore, that departure would depend on the arrival. If the arrival is late, the ensuing departure would also be late. A departure flight is created at a gate and consists of a flight that:
Just as an arrival can become a departure via the turnaround, the departure can generate an arrival that is dependent on the departure. An overflight is an arrival that does not land at the airport. The overflight traverses the airspace and exits the simulation.
Cloning of flights can be used to increase or decrease traffic on a route. Existing schedules can be modified and still reflect the original traffic pattern. This technique is helpful in estimating traffic congestion based on changes in scheduling.
Flights are identified by aircraft models. This, as well as other data, allows SIMMOD to distinguish different aircraft. They may then be assigned appropriate separation rules, sequencing, speeds, takeoff and landing characteristics, and size limitations. The aircraft is referenced by model number, airspace group number, and airfield group number. The aircraft model numbers are those as defined by the Integrated Noise Model (INM). The number defines size, including information for takeoff weights, fuel burn, and aircraft profile computations.
Airspace groups are divided into five basic types: Large/Heavy, Heavy, Large, Small, and General Aviation. The characteristics for each group include: speed ranges, holding queue, minimum separation, intrail separation multiplier, and wake turbulence sequencing. The airfield groups are divided into four basic types: Heavy, Large, Small, and General Aviation. The characteristics for each of these groups include: landing characteristics, takeoff characteristics, and gate occupancy characteristics.
The airspace is made up of an interrelated network of aircraft routes. The routes are defined by nodes and links. The aircraft move along these paths and are assigned a route as part of their input. The aircraft may not deviate from their assigned path.
The simulation does not check vertical and lateral separation. These aspects are only maintained as the user correctly (or incorrectly) defines the routes. A route is defined by a series of nodes connected by links in the direction of travel. The interaction of routes is monitored by the links and nodes they have in common. In other words, routes may cross, merge, or diverge.
Holding within airspace is done at a node. A specific flight pattern is not explicitly modeled, but the effects of the hold (time) are modeled as if the aircraft are waiting in a queue. The air route structure must be defined by the user to account for the holding pattern at a specific node. Link speed ranges are defined by maximum, minimum, and nominal speeds. Any speed within the range may be used by the simulation. The standard measure of speed is true airspeed in knots; however, inputs may be made using indicated airspeed in knots or Mach number and SIMMOD will make the conversion.
The effects of wind can vary by link by use of wind sets. Examples of wind sets are high versus low altitude links, physical location, or terminal approach versus en-route links. The simulation will include wind effects in all speed calculations and will effect time and fuel consumption figures.
A group of routes can be defined to handle different operations for an airport, making the route definition more complex than the basic node and link list. Each type of operation is a plan. Under a certain plan, various routes (e.g., a northern or southern flow) may or may not be available for use. As a plan changes, different routes become available.
Airspace logic considers the fundamental rules for aircraft movement on links, control at airspace nodes (holding), and three aircraft movement control strategies. An overview of this logic is discussed later in this report. The logic manages the simultaneous movement of aircraft on all routes. Each aircraft’s position is evaluated with respect to other aircraft when the aircraft is at a node. Based on data input, air traffic control decisions are resolved before the aircraft may move from the node to the link. A movement control strategy defines the logic used in controlling the aircraft. Each strategy or combination can have different effects on aircraft movement and can therefore be used to resolve different situations. The control strategies are applied when an aircraft enters a link and the link’s capacity has not been reached.
There are three levels of aircraft movement control strategies. From complex to basic, Level I is Node Arrival Control, Level II is Metering Control, and Level III is Flow Control. The more complex the strategy, the more complex the logic, and therefore, the more data input required. The three types of node arrival controls are QFIFO, SpeedFit, and MultiFit. QFIFO means that the first aircraft into the queue is the first out. It is the default control type for the simulation and recommended unless more control is required. SpeedFit allows aircraft to adjust speed, within range, when entering the node arrival queue in order to allow appropriate separation between aircraft. MultiFit attempts to make room for an aircraft at each position by adjusting the speeds of other aircraft in the queue, in addition to that of the entering aircraft.
Metering Control is an optional strategy that enhances the ability to control aircraft movement. Metering allows a look at downstream congestion by projecting aircraft positions at nodes along a route. The simulation will attempt to adjust aircraft spacing to minimize congestion or divert aircraft to a less congested route. The Flow Control strategy confines traffic by adjusting separation distances at nodes on the boundary of defined airspace. It is an optional control to prevent congestion problems that cannot be reasonably controlled by the previous two strategies.
Airfield structures use nodes and links to define:
Interface logic controls aircraft as they transition from airspace to airport or vice versa. This involves the coordination of flights, routes, and runways in the transition, creating a complex process. An interface node is defined to have special interface procedures associated with it, in addition to any other node characteristics it may have. These procedures define time and distance restrictions needed for a clear runway and for management of any associated runways.
Conditions that were initially set for a simulation may change during the simulation. This may be accounted for by resetting simulation parameters. SIMMOD allows such changes as using runways for takeoff or landing and corrections due to changing winds. The user can define the events and the times they are to occur. Some of the changes that can be made include:
There are several forms of output for results presentation. The report files can be exported, edited, and reformatted as desired. There are several data detail reports available. The four reports used for data interpretation are:
The data input echo report uses data from the application’s input files and reproduces it in an easily readable format. It is used to check the validity of the input data. The simulation log can provide information on every event processed. The analyst usually chooses specific information of interest for that report. It is used to verify and correct the air traffic control logic defined in the input data. The standard report provides global statistics on ground delay and travel time, air delay and travel time, and sector occupancy statistics. Other information in the report includes statistics on: total number of flights, aggregate flight delays, runway crossing delays, and random number seed values.
SIMMOD also includes presentation graphics and animation. The presentation graphics display charts of different variables over time. These include air and ground delays, route delays, delays experienced at individual nodes, and takeoffs and landings. The animation facility reruns the simulation on a screen showing aircraft move through airspace and over the ground network. It allows visual verification of structure and control data.
Integrated Noise Model (INM)
The FAA, Office of Environment and Energy, developed the Integrated Noise Model (INM) with assistance from ATAC Corporation, the Department of Transportation Volpe National Transportation System Center, and LeTech Incorporated. INM is widely used for evaluating aircraft noise impacts in the vicinity of an airport. It is used for FAR Part 150 noise compatibility planning and FAA Order 1050 environmental assessments and environmental impact statements. The system only requires a personal computer, 486DX, 66 MHz or higher, 32 Mb RAM, and Windows 95 to run. The computer program is written in the C++ programming language.
This section provides a general description of INM, which should be sufficient so that informed decisions can be made with regard to the model’s use in an aircraft routing and noise impact study. INM is an effective noise model and the standard methodology currently used by the FAA. It assesses changes in noise impact resulting from:
The INM aircraft profile and noise algorithms are based on SAE-AIR-1845 methodology. It is not a detailed acoustics model. This means INM does not account for temperature profiles, wind gradients, humidity effects, ground absorption, individual aircraft directivity patterns, and sound diffraction around terrain or buildings. It is an average value model that estimates long-term average effects using average annual input conditions. Therefore, predicted and measured values can and will differ. INM is not designed for single-event noise prediction.
A study is created by using a setup menu that aids in defining the input. The various functions to define are:
The latitude, longitude, and elevation can be entered manually, or an airport selection may be made from the INM list provided. If aircraft required are not included in the standard list, a list of substitutions is available. The substitutions provided are officially approved by the FAA for a noise study.
Thirteen noise metrics are available, including Day-Night average sound Level (DNL), the metric predominantly referenced throughout this project. The thirteen metrics are divided into two categories, A-weighted noise metrics and Tone-Corrected Perceived-noise metrics. Additionally, the user may define their own metric if that metric is not on the list.
Output graphics functions are used to view:
The output options also include the capability to display a list of:
An INM Technical Manual is available. It describes the flight-path methodology and the basic methodology used to compute a noise-level or time-above metric at a single, user specified observer, or at an evenly-spaced, regular grid of observers. The fundamental components for computing noise in this model are a flight path segment and an observer. For a given observer location, noise computations are performed on a flight segment. The methods used to compute the flight path segments and the methods for computing noise levels at an observer position are described in significant detail in the manual.
Airspace Design, Evaluation, And Planning Tools (ADEPT)
Metron, Inc. has developed an airspace planning and environmental impact assessment tool package: Airspace Design, Evaluation, and Planning Tools (ADEPT). The key tools within ADEPT are Airspace Design Tool (ADT), Noise Impact Routing System (NIRS), and Route Optimizer for Mitigation Analysis (ROMA). ADT is for development and analysis of aircraft routings based on current and planned operational scenarios. NIRS is for large scale modeling and analysis of noise impacts associated with current and future routings. ROMA provides two-dimensional and three-dimensional optimization of current or future routings to meet quantitative noise impact reduction goals. Other parts of ADEPT include Noise Complaint Analysis and Support (NCAS) and Sensitivity Analysis Tool (SAT). There will also be enhancements and variants of the above tools.
NIRS, in beta testing and owned by the FAA, is for use at multiple airports as compared to the Integrated Noise Model (INM) which is designed for individual airports. NIRS has been prototyped at Chicago and can, therefore, handle the much larger regional problem. NIRS also pinpoints causes of noise for possible mitigation. This leads to ROMA finding an improved track that reduces noise impact.
The important aspect of ADEPT is that it integrates all the modeling into one software package that performs the job of many. Previously, no single tool handled the entire task. ADEPT runs on a Pentium which translates to low overhead. The noise modeling, normally performed farther down the line, feeds back to help the process, creating a more dynamic tool than previously available. This tool interfaces with others on the market. This reduces if not eliminates one set of data being incompatible with another. However, the system requires data manipulation prior to input.
Total Airspace & Airport Modeller (TAAM)
The Total Airspace & Airport Modeller (TAAM) is an application for the simulation of airspace and airport operations. It is a gate-to-gate system that models the entire airside and airspace environment in detail, including pushback, runways, terminals, en-route and oceanic airspace. TAAM has been used in more than seventy major studies in the United States, Europe, and Asia Pacific.
The primary objectives of TAAM are:
TAAM allows evaluation of:
TAAM provides the capability to perform a "what-if" analysis. The real-world simulation can handle high volumes of traffic movements and provides three-dimensional, color graphics. The output can be formatted for input to noise modeling systems. There is a link for use in real-time support systems.
TAAM can provide applications related to the airline. An airline’s schedule can be analyzed to determine the impact of a delay. The global schedule can be analyzed and variations tested, leading to reduced delays and more realistic block times. Airline schedules, airport maintenance, or airspace changes can be analyzed from the hub operations perspective. The impact of severe weather, or a similar event, and its effect on that day’s schedule can be studied. The efficiency or proposed recovery measures can then be assessed. TAAM can model new regulations, separation requirements, aircraft type reclassification, curfew, or noise abatement procedures and their impact on airline operations can be assessed. The simulation can also model changes in fleet mix and upgrades to new aircraft and therefore allow the airline to assess the impact. Overflight charges can be calculated for international flights and alternate routes to avoid the charges may be evaluated.
TAAM provides airport related applications. It can perform airport capacity bottleneck analysis. TAAM can assist with studies in projected traffic growth. It facilitates determination of aircraft increases based on the airport’s present state. The model can handle new airport procedures for capacity constraints without disrupting current operations. The effects of future airport infrastructure components (e.g., new gates, taxies, runways) can be assessed taking into account projected traffic growth. TAAM aids in the evaluation of various airport extension proposals by comparing them via a measure of efficiency. TAAM is quite useful for new airport design which can make use of its advanced graphics and computer animation.
TAAM also has applications that focus on the en-route airspace. Its ability to assist in airspace redesign is considerable. This includes the analysis of oceanic airspace and its differences (e.g., reduced vertical and lateral separation). TAAM can assess the effects of new navigation technologies such as satellite navigation, GPS approaches, and Precision Runway Monitor equipment for parallel or crossing runways. TAAM also helps with special use airspace (restricted, military) that is only open during certain periods.
The Aircraft Activity Display System (AADS) is a real-time traffic monitoring tool. AADS receives flight plan data before departure and regular radar position updates of all airborne aircraft. It can also accept real-time weather data. Flow controllers can monitor all traffic in a large airspace or for certain categories of flights.
Four different software packages are described in this report: SIMMOD, the Integrated Noise Model (INM), Airspace Design Evaluation & Planning Tools (ADEPT) which contains the Noise Impact Routing System (NIRS), and the Total Airspace & Airport Modeller (TAAM). TAAM is for use by larger entities such as the FAA or the airlines. The others are more appropriate for smaller groups such as private consultants and citizen groups. The training required because of the complexity of TAAM is beyond that of SIMMOD or ADEPT.
However, all packages are considered effective and reliable provided the operators are well trained and the input valid. TAAM and SIMMOD data may be manipulated for use as input to INM to evaluate noise. Pending the beta-testing of NIRS and the FAA’s decision regarding future uses, INM is the present noise model of choice for a study.