Digital
Orthophotos: The Basis for Cost Sharing
By
Fred Sedgwick
Introduction
The digital orthophoto has a
variable scale and until a hard copy is made or it is
displayed on the GIS it has no scale. This variable scale
of digital orthophotos opens the door for cost sharing the
development of a common photobase. This ability to share
costs between public and private entities:
• Reduces the cost for all parties involved,
• Frees up funds for more frequent photobase updates,
• For taxing entities it makes more efficient use of tax
money,
• Facilities maps created using the common photobase may
easily be shared and
• Pooled resources allow purchasing higher quality
digital orthophotos.
This paper has three
sections-Finding a Common Base Map, Putting the Group
Together, and A Digital OrthoPhoto Case Study featuring
the Alamo Resource Conservation and Development Area, Inc.
(RC&D.)
Finding a Common Photobase
How was it done in the past
The key of a photobase is to see and interpret features on
the ground. These may range in size from a water shed to a
large ranch to a fire hydrant.
Differing needs for photo
interpretation require different scales. For instance the
Soil Conservation Service uses 1"=660' and
1"=1000' for its printed material, regional planners
commonly use 1"=2000' and the Appraisal Districts
commonly select 1"=400'.
To acquire the photobase
with these scales each entity has chosen to take aerial
photography at the best height to create the hard copy for
their photo interpretation needs. To illustrate the
differences of acquiring data at these scales, consider
the following table (The aerial extent is 1000 square
miles - a typical county in Texas):
Clearly the aerial
photography used by the Soil Conservation Service does not
satisfy the needs of the Appraisal District. Conversely
the photos needed by the Appraisal District can be too
large scale for the SCS.
The cost of imagery the old way
Cost is the key factor for any photobase acquisition. If
the final photobases are of similar quality the lowest
price will be used.
The following discuses the
cost to produce aerial photography for several entities at
their desired photo scale. The assumptions used are those
of the previous section and that the area of coverage is a
square of 1000 square miles.
Expanding on the previous
table we have the following table:
The range in cost is
$10,000 to $37,000 for the same geographical coverage at
different scales.
In a typical county there
are many taxing agencies and private businesses that need
to produce photobases similar to the above. The scale and
map production needs will be discussed in more detail
later.
A new method for acquiring common imagery
The digital orthophoto photobase allows data to be placed
in the GIS at the scale the user chooses. The vector data
is geocoded allowing the GIS to register maps (and other
feature data) over the photobase. The Appraisal District
can see the number of cars in the parking lot and the
swimming pool that has been added. The Soil Conservation
Service can see the farms, establish the boundaries of
estuaries, and study soil types. The Regional Planning
Commission can see major roads and locate impediments
(geographical features, man made things, etc.) to city
expansion. In essence the user of the GIS can view the
raster imagery, point and click and get the coordinates of
the feature (your house and the new pool). These can all
be tied to a data base that is capable of calls based on
geodetic coordinates.
Two terms are very
important Hardcopy and Softcopy, Hardcopy is the printed
map material, Softcopy is the imagery (or maps for that
matter) that is displayed by the GIS. Softcopy is digital
files that will be delivered on magnetic tape or CD-ROM.
Hardcopy can also refer to the aerial photography that is
created by photo enlargement as well as the printed
picture that comes from the GIS.
The USGS 7.5 minute Quad
sheet is currently the base map standard for mapping in
the United States. The USGS has defined a new product that
is called a Digital Ortho Quarter Quad (DOQQ) and it will
become as widely used as the 7.5 minute Quad. It will be
produced in Softcopy, it will have a ground pixel size of
1 meter, and it will have an accuracy of National Map
Accuracy Standards for a map scale of 1:12,000. The USGS
program to produce these across the U.S. is just getting
started and like the 7.5 minute Quad sheets (primarily
because of budget considerations) will take some time to
attain full coverage across the country. Estimates range
to the turn of the century. Because the DOQQ will be
produced by the USGS and will be available for public use
the cost will become very inexpensive. The USGS is
currently looking for joint sponsors to share part of the
cost and they will underwrite some of the cost. This leads
to a shared base map concept.
DOQQs have some specific
implication for use in GIS. DOQQs will be produced on
quarter quad boundaries so a naming and tiling convention
already exists. They will be produced with the same
coordinate system leading to consistency in mapping. The
registration, mapping, geodetic inventory and other GIS
activities are done on the computer system and the scale
of work is selected by the operator. A DOQQ can support
1"=400' scale very easily as well as any smaller
scale, for example 1"=660' and 1"=2000'. The
final printed material will probably not include the
raster imagery but will be vector maps. This can be done
at the scale needed by the user organization.
In the previous section the
cost was discussed to acquire a county photographic
backdrop GIS. For a similar county of 1000 square miles
the cost including new photography is near $50,000 for the
DOQQ product described above. The cost to each entity in
the common photobase is greatly reduced (depends on the
number of entities sharing cost).
Enabling Technologies
The world of computers is changing very fast. Faster, more
powerful computers provide the enabling technologies for
the creation of economical digital orthophotos. Three
computer related units are involved - computing speed,
random access memory and disc storage. There are some
software elements that are necessary (namely GIS) to
utilize these elements to their full potential.
The author (Sedgwick) has
for many years tracked the three elements that make
computers fast. For constant prices all three are changing
by a square factor. The author still remembers paying
$3,500 for 10 Megabytes of Disc storage, currently that
much money will purchase 10 Gigabytes. That is a 1000 to 1
increase in 15 years with no change in price. The
computing speed and memory size has taken similar leaps.
The increases in capability
have allowed the storage of large raster image files. This
has also provided the signal processing capability
necessary to create digital orthophotos and use them as a
backdrop for vector drawings.
You'll never get away from YOUR vectors
Raster Imagery works well for the base layer in a GIS
environment but when the time comes to show the boundary
lines of a piece of property there is no substitute for a
vector drawing.
For the Appraisal District the raster imagery serves a
number of purposes:
• Find taxable structures
• Help delineate property boundaries
• Arbitration hearings to show like properties
• Vegetation, crops and Forest.
For 9-1-1 mapping the raster imagery serves to:
• Find houses
• Map roads
• Set up the Emergency Service Number area.
For the Soil Conservation Service the raster imagery
serves to:
• Show cultivated areas
• Illustrate different crops
PUTTING THE GROUP TOGETHER
Is it appropriate to share costs
YES!!! In fact it is inappropriate to not share costs for
a common photobase. As previously discussed, if data is
shared the individual project participants can have better
data at less cost.
Governmental agencies and private
industry have the need for a high quality current
photobase. From the previous example the cost for a county
is approximately $50,000. If 15 participants share in the
acquisition of the data the cost to an individual
participant is only $3000.
This need for a good
photobase is shared at all levels - County, Regional,
State and National. Many private organizations have
similar needs.
Those that can share
The natural breakdown of organizations that can share in a
photobase project is done best by the geographical area as
discussed above - County, Regional, State and National.
The breakdown works well for both public and private
concerns.
Taxing, Law Enforcement and
Transportation all have organizations at all levels but
most have enough autonomy that they will make purchasing
decisions separately. The County Engineer working on
county roads will have a different budget than the Texas
Department of Transportation.
A Cellular Telephone
operator will have regional needs but might purchase the
single county where the next instillation is planned. This
company might also participate on a regional level.
Specifications that most entities can live with
Two ranges of scales often are discussed, Regional and
Large Scale. The scales are 1:24,000 to 1:4800 for the
regional scale and 1:2400 to 1:600 for the Large Scale. It
is generally conceded that the Regional Scale is needed by
almost all organizations, and that the coverage should be
of the whole region of interest. The larger scale is
needed by some organizations but is limited to smaller
areas.
The authors have done
extensive "non formal" research to discover
these scales, so they may not be perfect and the reader
may disagree. Even if the reader chooses to disagree, one
thing is for certain, there is a lot of entities in the
range of 1"=400' to 1"=2000' for their primary
mapping scales. The USGS went through a similar exercise
to determine the resolution and accuracy for the Digital
Ortho Quarter Quads. They chose 1 meter ground pixel size
and pixel placement accuracy associated with a map scale
of 1:12,000' (1"=1000'). It appears that this is a
good compromise.
A one meter pixel allows
display of the raster imagery at 1"=400' without
"seeing" the pixel, if inspected close enough
the pixel is visible. The authors have frequently seen
users view the one meter raster imagery on the GIS system
at scales of up to 1"=100'. If hard copy is produced
from the one meter digital imagery at scales greater that
1"=400' significant re-sampling will be required.
Pixel placement accuracy is
probably the least understood of the two basic
specifications pertaining to digital orthophotos. For the
USGS DOQQ program the pixel placement accuracy is
specified as having the pixel within its real position by
9 meters 90% of the time. If you use a GIS and place the
pointer on a feature shown on the digital raster imagery
and read the coordinates they will be within 9 meters of
the real position 90% of the time. The pixel placement
accuracy is largely dependent on the Digital Elevation
Model (DEM) used for ortho rectification. This process
also depends on the accuracy of ground control, number of
ground control points, aero triangulation and camera lens
radial distortion.
The specifications can be tabulated as follows:
• Photobase must be affordable.
• The 1:12,000 NMAS are affordable and do meet most
users needs.
• Black and White photography is more affordable than
Color and CIR.
• One meter pixel size.
• Where higher resolution is needed it can be ordered
for the individual area rather than the whole county.
• CD-ROM is the most useful media due to savings in
hardware costs.
The USGS has selected the
softcopy imagery discussed above as the next base map
product and as such will probably receive (in time) the
same popularity as the 7.5 minute quad sheets.
A CASE STUDY
Alamo RC&D
The Alamo Resource Conservation and Development Area, Inc.
(RC&D) is in the process of completing a shared cost
project in 6 counties around San Antonio, Texas. The Alamo
RC&D has taken on the following activities:
• Develop a set of specifications for the project.
• Select a vendor.
• Set up contracts with project participants.
• Manage the contract with the vendor.
• Distribute copies to all project participants in the
format for their GIS.
• Maintain an archive set of copies.
• Provide a method for non participants to purchase
coverage after the project is complete.
• Provide a path for updates in the future.
There are 26 RC&D areas
in Texas. Each is set up as a non profit corporation with
one person from each county serving on its board of
directors. The charter of the RC&Ds is to assist the
rural areas in attaining the services that the urban areas
have in place.
The process of building a
common base map from digital orthophotography has already
been started in the San Antonio area (Bexar County). This
project was the forerunner to the project discussed in
this paper.
In 1991 the Alamo Area
Council of Governments (AACOG in San Antonio) and the
Alamo Resource Conservation and Development Area, Inc.
selected a vendor and, as a pilot project, built digital
orthophotography for the County of Bexar. Entities in
Bexar County each provided a small amount of funds to
facilitate the project. The data purchased has been
distributed to the participating entities and most are
using it in their GIS activities. The knowledge learned on
this project is being used to do a better job on this
project.
Building on the Bexar
county project, Alamo RC&D moved forward with the
common photobase project soliciting individual funds from
organizations that have GIS needs in the 6 county area.
These six counties were selected because of their
proximity to the Bexar county project:
• Medina
• Bandera
• Kerr
• Hays
• Comal
• Guadalupe
One of the main reasons for
selecting these counties was that some of the entities in
the Bexar project asked for additional digital orthophotos
beyond the Bexar county line. The following discussion
highlights how organizations outside the individual
counties can be major contributors.
The steps in this project
are complete and waiting the final commitment of funds. A
vendor has been selected to produce the digital
orthophotos. The vendor has produced six quarter quads and
submitted them to the USGS for approval. As soon as USGS
approval is given the vendor will begin production.
About the Authors:
Fred Sedgwick is president of Sedgwick
Associates which specializes in customer driven unique
solutions to the development of land information. He may
be reached at 713-493-2584.
Wayne Griffin is the coordinator for the Alamo
Resource Conservation and Development. There are 26
RC&D areas in Texas. Each is set up as a non profit
organization with one person from each county serving on
the board of directors. The charter of the RC&Ds is to
assist the rural areas in attaining the services that the
urban areas have in place. Wayne may be reached at
210-426-5516.
Back
|