Geographic Visualization of ASDI Flight Plan Data David Hill GEOG 5561 The field of geographic visualization (or geovisualization ) has steadily developed and distinguished itself from standard scientific visualization as computing power has increased during the past two decades. We are currently witnessing radical developments in the ability of geographers, designers and data artists to render complex, multi dimensional spatial data sets in new, interesting, and scientifically valuable ways. My interest in the various methods used to visualize flight plan data began when I was exposed to Aaron Kolbin s Flight Patterns project. This project engaged me in a purely aesthetic sense, but also encouraged me to investigate how it was created, and what data the visualization represented. I discovered that it was derived from a spatial data set made available by the FAA, and that there have been many attempts at visualizing this data set in a variety of ways, for different purposes. A presentation of these flight plan visualizations illustrates the wide diversity and various functionalities of current visualization techniques. Visualization at its core is a method of structuring, formatting, converting, and displaying complex data sets. A large and complex amount of data can be difficult to understand in raw form. Often, data sets are compiled into spreadsheets and/or databases. Numerical data can be especially vexing unless the data is structured and formatted in a recognizable way. However, after the data is represented in a visual manner through histograms, scatterplots, or other visual forms, patterns and meaning can be discovered, usable information can be gleaned, and conclusions
about the data set as a whole can be made. The more complex the data set, the more important it is to represent that data as a recognizable visual form. Historically, one of the major drawbacks of data visualization is that detail was lost as the data was simplified and classified into visual form, so visualization was really best used as a way to give high level understanding of data sets. However, with modern computing power, it is more possible to let users drill down into high level data sets to glean specific details as needed. Interactivity is one of the hallmarks of effective visualizations now, and is one of the main components separating classic cartography from geovisualization. A fantastic example of the kind of multi dimensional data sets that can be understood using modern visualization techniques is the Aircraft Situational Display to Industry (ASDI) data stream made available by the FAA to companies in the aircraft industry. The FAA provides information about every civilian flight originating or terminating in the United States via ASDI. The stream is available in XML format, and includes data about each plane s position, altitude, airspeed, heading, destination, estimated time of arrival, and the tail number and carrier of the aircraft. The stream is updated every minute for every flight, giving it excellent temporal resolution. Since the feed is provided in XML format in real time, the data is easily parsed by companies interested in using the information to visualize realtime flight paths, anticipate late flight arrivals, or analyze flight patterns on a large scale. In my research, I found many different visualizations of the data.
One organization, Mobiata, used the information coupled with airport gate information to provide real time details about flight arrival and departures to mobile devices. In this use, Mobiata.com focuses on delivering www.mobiata.com low level, detailed information, not a small scale visualization based on overall patterns. While this sort of information can be useful for travelers, it is not very successful as a visualization of high level data. The University of Nebraska at Omaha s GIS Lab used historical ASDI data to create animations of individual flights. Their Animated Atlas is available on DVD and portrays the data as individual points (planes.) While a number of different scales are offered (continental, regional, local), the animation speed is fixed, and the data is portrayed as a series of frames, showing the latitude and longitude of each plane in the air. Altitude is not portrayed. The colorization of maps.unomaha.edu/animatedflightatlas individual planes generally classifies them by manufacturer or model. In any individual frame, we are unsure about the altitude or destination of the planes. The Animated Atlas does a good job of portraying the sheer volume of air traffic in the US, however. The animation is greatly sped up (about 1 hour/sec), so it is able to portray 24 hours of air traffic in less than 30
seconds. This is a high level portrayal of air traffic, but the viewer is not able to be more than a passive participant, limiting its usefulness as an interactive visualization tool. The FlightStats website also delivers real time ASDI data feed information, but adds powerful visualization tools. Users are able to access a wide variety of information, configure the way they view that information, and evaluate the information based on their own set of criteria. Information about the location of different airplanes is www.flightstats.com available on continental, regional, and local scales. Individual flight paths can be drawn as Google Map overlays using ASDI data and the data used to draw these paths can be viewed and easily downloaded. The FlightStats website was the most flexible and useful visualization of ASDI information that I found. It meets all of the criteria of effective modern visualizations. The website is well structured, renders a complex and dynamic data set, and is very interactive allowing the user to pick and choose what information to display. Most impressively, it provides details on demand, even when displaying information at small scales. Clicking on any individual plane will display complete, current ASDI data on that particular flight, including destination and arrival, current lat/long, altitude, airspeed, bearing, and more. One drawback is that FlightStats does not include archival information, and it
only portrays real time data. It does not allow the user to recall and play back archival information related to specific flights or timeframes. Aaron Kolbin took a different approach to portraying ASDI data in the creation of Flight Patterns in 2009. He was interested in portraying the data set, but he wanted to selectively filter out data that did not suit his needs, and emphasize the data that helped tell his story, which was the pattern of flight paths taken across North America. He used ASDI information from August 12 13, 2008. This included 205,514 flights. Each of these flights was sampled every minute to provide updated information, giving the data high temporal (and, as result, spatial) resolution. 26,552,304 data points were Aaron Kolbin's Flight Patterns processed to create Flight Patterns. The project is made available as static images (showing all flight paths over a 24 hour period,) animations (showing progression of flights across North America during daylight hours,) and as an interactive Google Map overlay, where users can select what sort of data they want portrayed (altitudes, aircraft manufacturers, or aircraft model) and then zoom in or out, or pan the selected view. One significant difference between Flight Patterns and the other visualizations listed is that the planes themselves are not represented, but rather, their paths are portrayed as colored lines. These lines correspond closely with the actual paths of
the individual planes. The lines themselves vary in color value depending on altitude, getting darker with higher altitude and lighter with lower altitude. Areas of light colors, therefore, tend to portray airports. As a result, Mr. Kolbin has created a unique and effective visualization. The Flight Patterns project is novel, providing a fresh look at the data, provides efficient and intuitive understanding of the problem, is aesthetically pleasing to draw attention to itself, and effectively represents the high level qualities of the data set in a compelling visual manner. Sources Dodge, Martin, Mary McDerby, and Martin Turner. Geographic Visualization: Concepts, Tools and Applications. Chichester, England: Wiley, 2008. Print. Dykes, J., Alan M. MacEachren, and M. J. Kraak. Exploring Geovisualization. Amsterdam: Elsevier, 2007. Print. Steele, Julia, and Noah P. N. Iliinsky. Beautiful Visualization. Beijing: O'Reilly, 2010. Print. "Aircraft Situation Display to Industry." Web. 29 Nov. 2010. http://www.fly.faa.gov/asdi/ "Animated Atlas of Flight Traffic over North America." Cartography and Geographic Information Systems Laboratory. Web. 29 Nov. 2010. <http://maps.unomaha.edu/animatedflightatlas/default.html>. "Flight Patterns." Aaron Koblin. Web. 29 Nov. 2010. <http://www.aaronkoblin.com/work/flightpatterns/>. Mobiata Mobile Travel Applications. Web. 29 Nov. 2010. <http://www.mobiata.com>. Track Flight Status, Airport Delays and Other Flight and Airport Information. Web. 29 Nov. 2010. <http://www.flightstats.com/go/home/home.do>.