Airborne: Texas Develops Hyperspectral Airborne Imaging Spectrometer
GIS is now providing cost-effective technology within the means of all governments in this country. Square in the center of it all is the digital orthophoto.
By David K. Nale

If your occupation involves managing a municipal or county government, the answer to the above question is more than likely "yes." Further, it most likely will be sooner than later! Geographic information systems (GIS) have, over the past five years, matured to the point of no longer being exclusive tools of the technologically adventurous or economically well off communities. Quantum leaps in cost efficiency and user friendliness have put GIS technology well within the means of all governments within this country. Indeed, it could be said that a local government not budgeting and planning for GIS is simply not being responsible to their citizens. Could you imagine saying that five years ago?
     What permits such bold statements? Several key advancements in GIS have transpired to result in its cost effectiveness.
• Dramatic advancements in computing technology. The cost/performance curve of using automated computing resources has gone off the chart. No longer are potential users budgeting for hundreds of thousands of dollars of computer hardware. Effective GIS computing can be implemented for very low thousands of dollars of investment.
• GIS software, now extremely powerful and multi-tasking has dropped to a price that budgeting isn't even an issue. Consider ESRI's ArcView 2, Intergraph's Vista Map, and MapInfo as examples. ¥ Global Positioning System (GPS) surveys are permitting accurate, real time collection of geographic features (road centerlines, utilities, etc.) in the field to be an every day reality. Airborne GPS has dramatically reduced the cost of geodetic control surveys (the mathematical foundation of any geographic information system).
• The municipal or county GIS map land base-the cost of which a few short years ago was measured in the millions-may now be implemented at figures approaching 1/5th of those costs.
      So the response to the question of whether GIS will be a reality in the governmental organization is no longer "maybe," but more likely "how soon?"
      The issues of map land base and data conversion are salient to both usability and cost of any geographic information system implementation. Square in the center of possible solutions is the digital orthophoto.
      Let's first take a look at what digital orthophotography is. A digital orthophoto is simply a scanned raster image, usually acquired from an aerial photograph, that has been accurately rectified with the aid of geodetic surveying and photogrammetry. The reason the technology is available is because today's computer hardware, software and storage capacity allow us to deal with the large volume of data created by the digital ortho process at a reasonable cost.
      A digital orthophoto will show visible natural and man-made details on the Earth's surface with extreme accuracy. Each pixel is associated with a precise geodetic coordinate, and measurement from one pixel to another can be an easy task. Of course, one limiting factor is pixel size. As the choice for pixel size changes from a six-foot pixel to a two-foot pixel to a six-inch pixel, the cost of the digital ortho process increases.
      Today digital orthophotos are produced from aerial photography. In the not too distant future, digital orthophotos in the 1 meter range of resolution will be produced from satellite imagery. The question now becomes one of whether the digital orthophoto is the most optimum solution to providing the geography in the soon to be acquired geographic information system. The answer is "to a certain extent." The digital orthophoto, dependent on resolution (pixel size), will provide a geographically accurate backdrop of the ground, and further depending on the cost/benefit ratio of producing the digital orthophoto, may well provide value beyongd that initial investment.
      Surface features are easily recognizable by the GIS user and the lay population in general. No one needs to explain to the viewer "I'm looking at my house and driveway and backyard." The visual interpretation that was required of lines on a traditional vector GIS land base is significantly reduced. Indeed, it is now a fun exercise for the average citizen to marvel at their own personal geography and the level of recognizable detail within the digital orthophoto. This is an important consideration. As the population at large is able to easily navigate through what was once complex GIS exercises, the political viability of the technology is assured (i.e., the viewer doesn't need to be a rocket scientist to understand that this is good stuff and it needs to be in place to effectively manage the community).
      The digital orthophoto by itself is not a complete GIS land base solution. It is however, a means of collecting and displaying a very significant amount of geography at a very reasonable cost. The cost performance is unquestionable. What is missing is the identification of those visible features on the digital orthophoto that the GIS software must "recognize" to operate. On the digital orthophoto image, a pixel is simply a pixel-no more, no less. Yes, it may have a State Plane Coordinate and Mean Sea Level elevation associated with it, but it has not been identified as a pixel within the image of a particular building or a pixel within the local landfill.
      By employing traditional photogrammetric procedures in combination with field acquired GPS survey control, ground features that must be identified for GIS applications can be collected at a very reasonable cost. For example, GIS applications may include emergency or trash collection, vehicle routing, utility management, tax collection or parcel management. To effectively use the information contained in the digital orthophoto image, the user must simply identify those ground features salient to the chosen GIS application(s) by photogrammetric and/or GPS methods, (dependent on cost effectiveness) and permit the digital ortho backdrop to display the balance of the community's geography. In other words, feature collection is application driven. Features that may need to be identified include:
¥ Street centerlines;
¥ Meandering driveways and buildings for emergency dispatch;
¥ Edge of pavement;
¥ Fences and walls for parcel location;
¥ Poles, manholes, catchbasins for utility management, and so on. The bottom line of this process is that the combination of multiple technologies, (digital orthophotos, photogrammetry, GPS) will result in a GIS land base that is affordable, application driven, and cost effective.
      Accepting the idea that digital orthophotos are cost effective, they are in the immediate near future and a well implemented GIS program can't be viable without them. Let's get down to the basics: "Which is best, black and white or color digital orthophotos," "What is the appropriate pixel size for the selected GIS application(s)," and "Should they be produced from aerial photography or space imagery, or simply purchased from outside sources such as digital orthophoto quarter quads (DOQQs) from the U.S. Geological Survey?"
      Consider the following:
      1. Color digital orthos are always more interpretable than black and white, but there is a cost both in file size and dollars. To produce a color digital orthophoto you must acquire color aerial photography (25 percent more expensive than black and white aerial photography), the photo must be scanned in red, blue, and green, thus tripling the scanning cost, image processing time in creating the digital orthophoto is increased due to tone matching between adjacent images and flight lines, and finally, disk storage of the color digital orthophoto will be three times larger than black and white of similar resolution. Total increase in cost of color digital orthophotos over black and white is in the range of 60 to 70 percent, but since color orthophotos are without a doubt more interpretable than black and white, the decision is one of serviceability.
      2. Resolution or pixel size of the digital orthophoto is application driven. In general, the GIS applications of a county would more than likely be satisfied by a pixel size corresponding to 2 to 3 feet on the ground. In contrast, most municipal or city GIS applications require a 6 inch to 1 foot pixel resolution.
      3. What should be the source of the digital orthophotos: aerial photography or satellites? The answer is "one or the other," or perhaps both, dependent on timing. The best guess in the industry for the availability of 1 meter resolution satellite imagery is the spring of 1998. The pricing and exact method of distribution of digital orthophotos produced from satellite imagery is still a matter of speculation. Until cost effective high resolution space born digital ortho images are a reality, airplanes will be the primary platform for photo or image five collection in support of digital ortho production. Further, airplanes will continue to be a primary platform of image acquisition for years to come. It is a matter of resolution, application, and cost.
      One meter spaceborne imagery may provide the geography for many county and state GIS applications. These images will create a "sea" of accurate land base data for the nation as a whole to draw upon, but within that "sea of consistent accuracy" will be a demand for "islands of more resolution and detail" than that available from space platforms, at a lower cost. There will obviously be a transition period over the next five years where appropriate sources of digital orthophoto land base data sets will change.
      Another source of digital orthophoto source data not to be ignored is the digital orthophoto quarter quad program (DOQQ) of the United States Geological Survey (USGS). The USGS is well on its way to compiling 1 meter resolution digital orthophotos produced from 1:40,000 scale National Aerial Photography Program (NAPP) aerial photos. It is the intention of the USGS to complete national coverage by the end of this decade.
      The DOQQs will be available on a local basis in both hard and soft copy. The cost for soft copy is estimated to be $32.00 per county.
      There are drawbacks-the DOQQs will only be as current as the NAPP photos from which they were created and the positional accuracy of each pixel will be in the range of approximately 50 feet (i.e., USGS quad accuracy). Additionally, the fundamental building block of the DOQQ is the 1 meter pixel. It is not a simple task to resample the data into the English foot of the State Plane Coordinate System.
      Is there an orthophoto solution in the future of effective GIS management? If it's the management of a city or county government, it is a pretty safe bet the answer is "yes." Whether it will be color or black and white, 6 inch or 1 meter pixel resolution, sourced from aerial photography or space platforms, are all points of important discussion and consideration.
      A qualified GIS consultant can advise you of cost performance issues related to the above concerns. As you consider these issues keep two things in mind: GIS is forever and change in technology is constant. GIS implementation will continue to be more cost effective and it will be hard to remember that cities and counties were ever managed without it.

About the Author:
David K. Nale is a certified photogrammetrist, ASPRS, certified mapping scientist, GIS/LIS and president of Aerial Data Reduction Associates Inc. He served with the U.S.A.F. and others as a photo intelligence officer and is recognized as an expert in photo interpretation. As chief photogrammetrist, he has managed over 2,000 photogrammetric and GIS projects across the nation and abroad. Nale is the first person ever to be titled certified mapping scientist, GIS/LIS by the American Society of Photogrammetry and Remote Sensing and serves on the selection board for the society.

Back