Locating new or relocating existing transportation
facilities requires terrain information that helps planners design and
evaluate alignments and grades for alternative corridor locations.
Collecting such information through ground surveys and photogrammetric
mapping is accurate, but also time consuming and expensive, and recent
research suggests that often only the final design stages of a corridor
project require such accuracy. New surface mapping methods, such as
Light Detection and Ranging (LiDAR) imagery and digital aerial
photography, may instead be used for a project’s planning stages,
allowing more targeted photogrammetric data collection for final
corridor alignment. The latter process could be limited to a much
smaller area, reducing time and cost.
To help quantify the economic rationale to employ remote sensing and
related geographic information system (GIS) technology for
transportation planning, the National Consortium for Remote Sensing in
Transportation-Environment (NCRST-E) estimated time and cost reductions
associated with remote sensing techniques compared with traditional
ground surveys and photogrammetric mapping. Here are the results of
three projects: Washington’s I-405 corridor, Iowa’s Highway 1 corridor
and North Carolina’s Highway 311 corridor. Due to the varied nature of
such projects, they don’t lend themselves to a priori standardization to
provide a rigorous analysis of remote sensing benefits. Thus, this is a
retrospective analysis of the value of remote sensing and GIS for the
three projects.
Washington’s I-405 Corridor
Using the I-405 corridor in Washington state as a test case, several
agencies and organizations collaborated to demonstrate and assess
the applicability of commercial remote sensing products and spatial
information technologies to environmental analysis in transportation
planning. Support was provided by the Washington State Department of
Transportation (WSDOT), Oak Ridge National Laboratory, ERDAS Inc.,
Space Imaging, the U.S. Environmental Protection Agency, Wisconsin
DOT and the Puget Sound Regional Council.
Methodology
The Washington case study involved more than 150 smaller projects
and used a post-project survey to determine cost and time savings
associated with remote sensing. The case study divided the
environmental planning studies into 11 “disciplines,” focusing much
of its economic analysis on determining the benefits remote sensing
delivered to each discipline. The study not only addressed possible
reductions in time and costs associated with remote sensing
technology, but also attempted to evaluate the value of the remote
sensing data.
A difficulty in estimating remote sensing
costs for a project organized by disciplines is that costs of remote
sensing data once collected may be applied to assessments performed
for more than one discipline. Typically, much of the land use and
land cover (LULC) classification is done simultaneously for
different disciplines. Because costs weren’t collected separately
for each discipline, an average cost per discipline was estimated by
dividing by the number of disciplines in which LULC maps were used.
The costs of developing, compiling and presenting the information,
data and maps in the environmental impact statement (EIS) were
obtained from the contractor team responsible for the different
disciplines. Time estimates to complete the remote sensing/GIS and
EIS products were based on actual calendar time, not person months
or full-time equivalent time calculations.
The value of the remote sensing/GIS products was estimated by
surveying users. It is important to note that the increased value
discussed in this case study result from a survey of 13 respondents,
and not from remote sensing experts. Some respondents provided
answers for disciplines separately, and others answered for all
disciplines in general.
Results Cost Comparisons: Table 1 lists the costs of completing the
EIS for each of the 11 disciplines. The average cost of producing
the remote sensing/GIS products was about $6,000 per environmental
discipline. With remotely sensed imagery, data processing time and
costs, as well as the time and costs of producing maps and related
spatial statistics, increase marginally with the number of
environmental disciplines, i.e., the total time and cost to develop
remote sensing/GIS products for 11 disciplines isn’t much greater
than the time and cost for one discipline. The estimated time to
complete the products was eight months for all products and
disciplines.
The cost per discipline ranges from $31,900
for floodplains to $243,300 for transportation by traditional
methods. Thus, the average cost of the remote sensing/GIS products
ranges from 2.5 percent to 19 percent of the cost to complete work
on an environmental discipline. Overall, the cost of the remote
sensing/GIS products for the 11 disciplines was $66,400, compared
with the $844,000 estimated to complete these 11 disciplines by
traditional methods. The greater the number of environmental
disciplines, the more cost-effective this type of analysis can be.
Time Comparisons: An EIS using conventional methods is expected to
take about two years, based on the discipline taking the longest
time, fish and aquatic habitat. The estimated time to complete the
work on all of the remote sensing/GIS products was eight months.
Only three of the disciplines—farmlands, floodplains and
recreational areas—were completed more quickly with traditional
methods.
Value: Respondents’ assessments of the value of remote sensing/GIS
products are summarized in Table 2. Fish and aquatic habitat, land
use, upland vegetation, habitat and wildlife, and wetlands assessed
remote sensing/GIS products to be most valuable. Respondents from
these disciplines suggested that a relatively high percentage of the
work done for the EIS could be achieved using remote sensing/GIS,
and the value of the remote sensing/GIS products equals or exceeds
its cost. Respondents from the farmland and floodplains disciplines
judged the technology to be somewhat less valuable. It represented a
smaller part of the overall EIS work, and its value was less than
its cost. The remote sensing/GIS products were least valuable for
the environmental justice, recreational resources and transportation
disciplines.
Iowa’s Highway 1 Corridor
Iowa’s 18-mile-long Highway 1 corridor through Solon, Iowa, is a
two-lane, undivided state highway with north-south orientation. The
corridor was selected from existing DOT projects based on the
existence of surface elevation photogrammetric data and the lack of
significant changes within the study area since photogrammetry data
were completed.
Photogrammetric data were available from the Iowa DOT for a
10-square-mile area around the corridor. The study segment begins at
the Interstate 80 interchange near Iowa City and ends at the
junction with U.S Highway 30 near Mount Vernon. The highway passes
through the town of Solon, the location of a proposed bypass, at
about the corridor’s midpoint. The corridor is characterized by a
variety of terrain: rolling farmland, the small town of Solon and
significant elevation changes at the Cedar River.
Methodology
The Iowa case study compared two different corridors, Highway 1 and
US 30, to compare the use of LiDAR in conjunction with
photogrammetry vs. the use of photogrammetry alone. One site used
the combined method and the other used the traditional. Differences
in costs and time were used as an estimate of the advantages of
using remote sensing techniques.
The times and costs for the US 30 corridor were estimated potential
savings, whereas the time savings reported for the Highway 1
corridor were collected as part of the project. For the Highway 1
corridor, traditional and combined data collection methods were
compared to determine whether the use of LiDAR would result in more
rapid data collection, production and delivery than photogrammetry
alone. The latter work had been completed for the corridor prior to
the NCRST-E research project, which addressed the utility of LiDAR
in the project.
Results
Cost estimates for US 30 are followed by the same estimates for
Highway 1, then the differences are noted and discussed. US 30 was
used as a baseline for comparison.
U.S Highway 30: The time required using only photogrammetry for the
46-mile corridor was estimated to be approximately two years. In
comparison, the combined method, using remote sensing and
photogrammetry, required only 13 months. The combined method
required five months for LiDAR data collection and analysis for
preliminary corridor siting and eight months for photogrammetry to
map the final alignment. In this case, 11 months were saved using
the combined method compared with the traditional method. In terms
of cost, using traditional photogrammetric mapping alone for U.S. 30
cost $500,000. When combined with LiDAR, the photogrammetric mapping
costs dropped to $100,000, with only an additional $150,000 required
for the LiDAR component, thereby reducing the overall costs by 50
percent.
Highway 1: The traditional method using only photogrammetry required
2,670 hours, and the LiDAR method required 598 hours. The resulting
time reduction using LiDAR was 2,072 hours, or approximately 450
percent. However, this comparison doesn’t include the additional
cost for photogrammetry to obtain final alignment as was reported
above for US 30.
North Carolina’s Highway 311 Corridor
North Carolina’s 15-mile Highway 311 corridor connects I-85 to US
220 in Randolph County near High Point, N.C. This is a rapidly
urbanizing area with potentially substantial environmental impacts
from transportation construction. This study area was selected based
on substantial wetlands that could be protected with more
sophisticated planning and construction techniques.
Methodology
The North Carolina Department of Transportation (NCDOT) evaluated
the accuracy and applicability of the LiDAR data for transportation
planning and design. The North Carolina case study focused its
efforts on obtaining cost savings estimates for specific excavation
costs—that is, the difference in excavation costs associated with
data obtained from remote sensing vs. traditional methods.
Results
NCDOT determined LiDAR data were sufficiently accurate and readily
applicable to preliminary design activities associated with
transportation projects. LiDAR data, obtained as part of the NCRST-E
project, supported mapping for preliminary design activities. NCDOT
transportation projects traditionally use mapping at 1:1,200 and
1:2,400 horizontal scales with 2-, 4- or 5-foot contours; LiDAR data
met these needs in the preliminary design phase.
The North Carolina Flood Plain Mapping Program (NCFMP) has LiDAR
data for approximately 80 percent of the state. Preliminary design
mapping (1:2,400), digital terrain modeling, orthophoto
rectification and preliminary earthwork calculations are all
preliminary design activities for which NCDOT regularly uses LiDAR
data.
The NCFMP LiDAR is reviewed and edited with 3-D stereo imagery, and
photogrammetric break lines are collected at significant features.
Although the NCDOT Photogrammetry Unit hasn’t formally documented a
cost or timesavings using existing NCFMP LiDAR data, project
experience suggests that it reduces the time to generate digital
terrain models by approximately 30 percent compared with a
photogrammetric technique.
Final Results
All three projects indicate that using LiDAR data can expedite the
planning and siting of transportation corridors. Designers can begin
preliminary analysis much sooner with LiDAR data, as environmental
conditions—e.g., sun angle, leaf off and cloud cover—won’t prolong
the process to the degree associated with photogrammetry. Aerial
imagery and LiDAR data can be collected concurrently, thus reducing
total acquisition time in terms of flight hours. The increased
availability of data means that terrain can be analyzed earlier in
the siting process, allowing potential problems and issues to be
identified and addressed more quickly.
Although LiDAR data have an advantage early in the siting process,
photogrammetry is favored when more intensive data analysis is
needed to define final corridor alignment. LiDAR data can produce
time and cost savings through more expedient large-scale data
collection; more costly, time-consuming methods only would be
necessary for limited areas.
