Airport Zoning Regulations :  Using Geographic Information Systems (GIS) in analysing impacts upon surrounding communities

© A.Krisciunas

Powerful analytical tools offered by Geographic Information Systems can extend the reach of business geomatics well beyond mapping, in serving clients.  While high-quality maps offer clients unprecedented visualisation of their properties (“a picture is worth a thousand words”), and provide them insights not even evident in such non-visual reporting formats as spreadsheets or bar-graphs, mapping represents only the tip of the iceberg for what a business geomatics firm is capable of offering its clients.

In particular, MapInfo facilitates location-based decision-making.  Clients may pose “what-if” scenarios based on questions about what surrounds their properties, or what is between their adjacent properties or the properties of other government departments and stakeholders.
 


 
Part I :  Case-Study, Winnipeg International Airport and Surrounding Residential Development

Federal governments serve airports in many roles, including, the provision of zoning advisory services.  Even with de-regulation of airports throughout the industrialised world, the need is still there for a centralised approach that elicits standards amongst the airport partners, particularly in light of the potential for significant real estate related liabilities resulting from increasing air traffic.

Canada's airports are facing rapid change, even in the near future.  They include, among others :

In responding to these changes by advising municipalities on their zoning by-laws, one set of questions that can be asked centre on residential land-use.  It may be important, for instance, to know the assessed value of homes in surrounding areas, or the rate of residential growth, if any, under flight paths.

We can do this quickly, efficiently and with minimal error if we assemble our data and work with it entirely within a GIS.  That is, although the data may originate from spreadsheets or “flat-file” databases, we bring them all together under a GIS and use the map as the basic “workbench” on which we do the job.

We construct the “toolbox” by layering the following information :

The following map illustrates what this toolbox looks like at the national level.  The 4 components listed above become “layers” (an important concept in GIS).  All layers exist at all scales, but at this scale, the Demographic data layer is too small to be visible.  However, this is where we begin the process of drilling down to Winnipeg International Airport :
An exciting enhancement to recent versions of MapInfo is the “HotLink” feature[1].  On the above map, it’s indicated by the blue and teal icons.
 
In this particular case, the “Assets” layer provides a hotlink directly to the DFRP webpage of the given airport, using the DFRP Property Number (PN).  A Netscape or IE browser is automatically launched and the asset's DFRP record appears, while the MapInfo session continues in the background.  Click the Winnipeg International Airport icon, for instance, and the DFRP Treasury Board record for DFRP #12738 pops up.
 
The “Aerial Photographs” layer uses its hotlink to launch a web browser and bring up the air-photo at that particular latitude/longitude from the database of the Canada Centre for Remote Sensing (CCRS).  The following table indicates the icons used to display the “hotlinks” on the MapInfo coverage:
DFRP ( “jet-plane” ), linking to :  http://www.tbs-sct.gc.ca/dfrp-rbif/
GCDB ( “radar-dish” ), linking to :  http://lambert.ccrs.nrcan.gc.ca/

Now, let’s begin the drill-down.  The following image shows a map of west-central metro Winnipeg (scale not shown, only for context).  As the number of “radar-dish” icons illustrate, the air-photo coverage around the airport is quite good.  (As a side-note, the air-photo of the runways of Winnipeg International Airport on the image below was not taken from CCRS but, surprisingly from DFRP!  It was saved from DFRP as a .JPG, then opened as a raster image in MapInfo and registered to 5 exterior points with latitude/longitude in decimal degrees to 4 digits accuracy, using the “DFRP Info Tool” to obtain the latitude/longitude) :

 
Further “drilling-down” allows us to bring up the demographic layer.  This then permits some important visualisation of data on the basis of geography that would never be apparent by simply looking at spreadsheets or tables in a report.  For example, the following map illustrates the density of residential dwellings :
(Note:  The very small areas, or “polygons”, that are red (“over 49,000”) are invariably high-rise apartment buildings when this map is reproduced for most Canadian cities.  In fact, the only polygons that are red on this map are the extremely small ones; all turn out to be high-rises when the MapInfo data are examined in detail.  The nature of this type of mapping (called “thematic”) is such that one must sometimes deal with a “quantum-leap” when classes are defined – in this example, the vast majority of polygons are medium to low density, while a small number are extremely high density.  Demographic data often pose this kind of problem.  Decisions made on final thematic maps usually involve several rounds of “trial-and-error” to arrive at an image that adequately projects the nature and “feel” of the neighbourhood, without masking important extremes).
 
Airport zoning regulations serve two major clients :
The following 2 “sub-cases” use thematic mapping to analyse the nature of residential development around Winnipeg's airport, and some GIS “functionality” to perform simple geographic-based calculations, to provide some estimate of impacts of the airport.
 
 
Sub-Case 1 :  Impact on Residential Property Values
 
Estimates of property value (from the Census) can be a “proxy” of assessed value, to determine impacts upon the tax resource of a city.  Municipalities are concerned over potential losses in assessed value.  First and foremost, the municipality must protect the equity that individual taxpayers have in their homes.  But lowered assessed values also result in a subsequent loss of property tax revenue; the concern of the municipality then becomes that of collective loss and not just individual loss.  Thus, there are two good reasons for measuring potential losses in assessed value.
 
Non-residential land uses can have significant negative impacts upon surrounding residential land use.  Worse yet, some non-residential uses, such as factories, may start off with little impact, but introduce a new process that begins to worsen conditions in surrounding residential areas.  An airport is not unlike a factory in this sense, in that decades ago it may have had little negative impact upon the surrounding area, but conflicts today with homes and businesses nearby.  An example of this is Thunder Bay International Airport.
 
In the extreme, a municipality may be forced to determine additional levies for noxious non-residential land-uses.  If a factory, for instance, were to increase certain negative activities (e.g.: air pollution, traffic), the resulting drop in surrounding residential property values, and, thus, assessment, can be converted to an estimate of lost municipal tax revenue.  Depending on enabling legislation, this amount could be charged against the factory's tax bill as an offsetting additional levy.
 
To be fair, in the case of airports, there can also be considerable positive impacts upon surrounding areas, that can increase assessed value.  If flights are added to hitherto unavailable destinations, or flight frequency is increased, this can be a bonus for some industries.  It could also be a good reason to re-locate that much closer to the given airport.  In high-technology centres like Ottawa and Calgary, this is precisely what is now happening: increased flights from Calgary to Phoenix, or Ottawa to Boston, for instance, are very appealing to local hi-tech firms, and in the case of both airports, developers with land near the respective airports are touting their advantage in being minutes away form the airport.  As well, airports all over Canada have always been attractive to major hotels (an attraction not necessarily shared by factories!).  GIS, again, can be a highly effective tool in analysing this.
 
However, for the purposes of Sub-Case 1, we’ll consider the negative effects, and assume that noise and accident risk are the major contributors to a loss of assessed value in residential areas near Winnipeg’s airport.  Thus, a we wish to focus on those neighbourhoods directly under flight paths.  The following map shows value of owned dwellings near Winnipeg’s airport (1996 Census) :
 
The next step, use of geographic “operators” to estimate actual impact, involves the use of “buffers”[2].  We begin by adding “vectors” to the map to represent invisible extensions of the runways.  Aircraft are either taking-off or landing along these vectors; they are, essentially, flight paths.  Next, we use MapInfo to create the buffer as an artificial geographic feature surrounding the vectors, based on information provided by the user.
 
MapInfo has the capability to establish complex buffers.  For instance, the buffer radius may be taken from a data field associated with other geographic features, or even calculated on the basis of an algebraic expression.  The buffer radius could, for instance, be a function of the average size of aircraft using the runways, the presumption being that larger craft have a wider noise impact on the ground.  For this case-study, however, a simple 1-km radius is used to create a single buffer based on the flight paths, as vector extensions of the 2 runways.  The following maps show how this buffer is constructed :
The blue squares represent the estimated flight path vectors for Winnipeg International Airport, manually created[3]:
The yellow zone is the 1-km “buffer” that is automatically created by MapInfo :

A “geographic query” is then made in MapInfo to select those EAs that “intersect” (another important concept in GIS) this buffer, and statistics are calculated for the selected EAs :
 

Selected EAs :
Statistical summary :

The results show an average value of ~$68,000 for the homes within a kilometre of the flight paths of Winnipeg’s airport, and that almost 60,000 people live in this zone.
 
 

Sub-Case 2 :  Encroaching Residential Land Use

From the airport's point of view, there may be less concern with property value, and more concern with growth of nearby residential land-uses.  If the municipality does not sufficiently play its part in accommodating present and future needs of the airport, too much residential growth may occur in nearby areas that will stymie the future options for the airport.  For Winnipeg’s airport, an important question, then, may be to ask where recent residential development has been occurring in the west end of the metro area, as this may be a harbinger of where it will occur in the future.

This can be determined from the 1996 Census, based on the number of dwellings constructed in a certain decade.  In the following map, each dot represents 1 dwelling constructed since 1980, based on EAs :

 
South of Ness Avenue would appear to have been the best area for residential development[4], since it is “nested” in the shadow of the two vectors, away from the flight paths, with less exposure to aircraft.  However, that's not where the housing has been built recently, as the above thematic map indicates.  Of course, there are many factors that influence the location of new housing, but one would assume that homeowners, making rational choices, would place a high level of importance on the influence of a nearby airport, and that homebuilders would respond accordingly.
 
Yet this doesn't seem to have been the case – this map shows that some of the highest volume of new residential construction has been occurring south of Saskatchewan Avenue, directly under the western vector of the airport!  Using the same type of geographic query as in Sub-Case 1, we find that 1,705 new homes have been built in the 1-km buffer zone in the past 20 years.
 


 
Part II :  Extension of the Winnipeg International Airport Case-Study to Nation-wide GIS analysis
 
Obviously, the analysis done in Part I for Winnipeg’s airport can be extended to all airports.  As stated above, GIS can do this quickly, efficiently and with minimal error.  For example, doing a “quick-and-dirty” exercise with 1996 Census data, using the above GIS methods, reveals that, within 5 kilometres of all 75 airports listed in DFRP :
To obtain these 4 items of data, the buffers were created, CTs selected, and statistics calculated in less than 15 minutes!
 
Limitations of time required that census tract (CT) rather than enumeration area (EA) data be used for this exercise on the 75 airports across Canada listed in DFRP, and this brings in a caution on how geographic data-sets are used in GIS analysis.  Basically, there is a trade-off: analysis is quicker with smaller data-sets, but smaller data-sets result in “coarser” analysis.  To arrive at the above results, we could analyse 30,000 EAs, or 4,500 CTs.  Obviously the latter involve faster processing, but in doing so, we could seriously mask differences within a census tract that would have been brought out by doing it at the enumeration area level (the average CT has 5-10 EAs).
 
The Canada-wide exercise using CTs resulted in the following typical selections … :
… which illustrate the problem of “coarseness” (e.g.: Pearson Int’l, west side).  While working at a “finer” level, such as the EA, is always preferable, the limitations of working at a coarser level can be somewhat overcome, if time or expense is of the essence.  For example, since census tracts are quite large and sometimes irregularly shaped, it may be inappropriate to use a simple “intersects” function.
 
For example, consider this particular buffer-selection from the actual Winnipeg airport query in Sub-Case 1 :
Using “intersect”, the buffer for the south-east flight path vector picked up EA# 4601230, the bulk of which lies outside the buffer.  Thus, the “intersect” option alone may pick up areas not appropriate to the study. 

(NOTE:  Those familiar with Winnipeg will note this is the Winnipeg Zoo area, and the smaller part of the EA under the buffer is, in fact, where most residents live.  In this case, then, the inclusion of the whole of EA# 4601230 turns out to be appropriate.  This is rare.)

MapInfo allows more sophisticated forms of geographic processing to help alleviate possible inaccuracies.  For example, to get more accurate data at the Census Tract level, we would “slice” portions of the census tracts corresponding only to the buffer; the buffer becomes a “cookie-cutter”, and the demographic data values for the cut-out portion are pro-rated down according to the measured area of the cut-out portion.  Thus, there is virtually no end to the sophistication that a GIS such as MapInfo affords in geographic analysis.
 


 
CONCLUSION

There is an exciting opportunity here to go well beyond mapping in assisting airport zoning regulations initiatives.  Using tools such as buffers, geographic operators and processing, and map algebra, we can provide the client(s) with analysis more closely reflecting the reality on the ground, of the interaction between airports and their municipal partners.  Furthermore, with high-end, user-friendly, low-cost applications, such as ProViewer, we can put, in the hands of the same clients, useful tools to empower themselves with GIS analysis … without them having to be experts.


 
 
 

APPENDIX – Example of Hotlinking
 
In the following 2 images illustrate the “hotlink” feature of MapInfo, referred to above.  An area in south-east Orleans (Ontario, Canada) was chosen to give a “local” reference.  The yellow dot on the air-photo corresponds to the “radar-dish” icon on the MapInfo map.  Note the curve of Trim Road on the map and on the air-photo, as a reference.  The air-photo is dated 1979, so the urban development in the “10th Line” area is not apparent.  More up-to-date aerial or satellite photo coverage could be made available for this project, however, through collaboration with the Canada Centre for Remote Sensing (http://www.ccrs.nrcan.gc.ca/).
Below is the map (east/west extent = 5 km) from MapInfo, using the same layers as the Winnipeg International Airport maps in Part I, that sources the air-photo image to the right.  Note the “radar-dish” icons in green, used to “hotlink” to the CCRS website to bring up the air-photo; the icon on Trim Road corresponds to the yellow marker on the air-photo image : 

An overview of the present air-photo coverage of CCRS (~37,400 points) is illustrated in the following map of Canada :

 
 
 
 

[1]  By clicking on a map feature in a particular layer, one can launch any number of separate tools, even ones outside the MapInfo session itself.  In this case, the tool is a Netscape browser, bringing up websites related to the individual geographic features.  An added bonus, with the latest release of MapInfo, is that this hotlink feature is also available for the ProViewer software that MapInfo offers free-of-charge (http://www.mapinfo.com/proviewer/), so that non-experts can easily view GIS information.  In this case, sophisticated coverage around airports can be assembled and provided transparently to a client, with ProViewer, allowing them to :
  • query the DFRP data on individual sites,
  • view air-photos of land surrounding airports, and
  • make simple calculations of demographic data for neighbourhoods around airports,
all with no need for any GIS training.  Please see :  APPENDIX – Example of Hotlink.
 
[2]  GIS uses the term “buffer” in a somewhat unconventional way.  To a chemist, a “buffer” is a solution that is able to maintain a pH level despite the introduction of acids or bases; to an urban planner, a “buffer”, similarly, maintains a neighbourhood's tranquility despite the introduction of of disruptive land uses (eg: factories or highways) nearby.  A GIS “buffer”, however, is simply a polygon that surrounds another geographic shape, usually conforming to that shape; it does not have an intrinsic functionality as would the “buffer” of a chemist or urban planner.  It does, however, lend itself to geographic processing.
(NOTE:  This author disagrees with the GIS industry's use of the term "buffer" because of its historical connotation, as illustrated in the above examples taken from such diverse fields as chemistry and urban planning, and feels that a more appropriate term should be derived ... but that's another story!)
 
[3]  The air-photo depicts a 3rd north-south runway at Winnipeg’s airport.  However, this runway is likely used exclusively by CFB Winnipeg, and is not open to normal civilian aircraft traffic, except perhaps for emergencies.  As well, even if it were open to commercial air traffic, most large commercial aircraft in Western and Central Canada will invariably use east-west oriented runways, as these coincide with the direction of the prevailing winds.  Thus, this north-south runway is ignored for the purposes of Part I.
[4]  Notwithstanding the north-south runway described in the previous footnote.  Since we do not have additional information on it at this time, we must assume it generates less air traffic and, thus, is less disruptive on Ness Avenue.
1