using lidar to study surface water runoff and impervious surface

USING LiDAR TO STUDY SURFACE WATER RUN-OFF
AND IMPERVIOUS SURFACE DELINEATION
Thomas L. Pagh
ASPRS Certified Photogrammetrist
Business Development Manager
i-TEN Associates, Inc.
2548 SE Ankeny St.
Portland, OR 97214
[email protected]
Carol Murdock
Technical Services Specialist
Clackamas County Water Environment Services
9101 SE Sunnybrook Blvd #441
Clackamas, OR 97015
[email protected]
Brenda R. King
President / Owner
Pinnacle Mapping Technologies, Inc.
8021 Knue Road, #113
Indianapolis, IN 46250
[email protected]
ABSTRACT
As part of a Surface Water Master Planning effort, the Department of Water Environment Services, (WES),
Clackamas County, Oregon, acquired LiDAR data covering a 52 square mile area in northern Clackamas County.
Project deliverables included raw LiDAR data, 1' contours, color orthophotos with 6" pixel resolution, plus bare
earth and canopy terrain data. Subsequently, WES contracted with i-TEN Associates, Inc., Portland, Oregon, to
create impervious surface polygons using the existing LiDAR data and orthophotos. This paper describes how
WES is using the processed LiDAR data in their Surface Water Master Planning process. Also, the methodology
used to create highly detailed and accurate impervious surface polygons, using the LiDAR intensity data and 6”
pixel, color orthophotos, is described.
SURFACE WATER MASTER PLANNING EFFORT
The purpose of this study was to estimate the amount of impervious surface that could potentially be created by
urbanization in the Damascus Urban Growth Boundary Expansion area in northern Clackamas County, Oregon.
Clackamas County - WES provides sanitary sewer and surface water management to the unincorporated portion of
northern Clackamas County and several cities in the general vicinity.
In 2004, the METRO Regional Planning Council for the Portland metropolitan area added approximately
1,900 acres of land in northern Clackamas County to the regions’ Urban Growth Boundary (UGB). This area was
formally called the Damascus UGB Expansion Area and has since been incorporated into the City of Damascus.
WES will presumably be the provider of sanitary sewer and surface water management for the City of Damascus
and, therefore, will need to develop both a sanitary sewer and a surface water master plan to provide these services
to the developing area.
In March of 2004, Clackamas County – WES obtained new LiDAR data and high-resolution orthophotography
for the study area (Figure 1). These data were then used to develop one-foot contours, six-inch pixel resolution
orthophotography, and a high-resolution Digital Elevation Model (DEM). These data have since been used to
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
develop hydraulic and hydrologic modeling in support of the surface water master planning effort and have been
instrumental in the development of a sanitary sewer trunk system for the area.
Figure 1. LiDAR Project Area.
IMPERVIOUS SURFACES
As the agency in charge of surface water management, WES is responsible for maintaining the water quality
and overall health of the urban streams within the service area. Unlike many urban areas in the United States,
northern Clackamas County has maintained a large portion of its small urban streams as aboveground open channel
systems. A strong desire on the part of the public to develop livable communities, combined with strict
environmental regulations, provides further incentive for WES to maintain the health of these systems in the face of
urbanization. One of the biggest challenges to maintaining stream health is the amount of impervious surface that is
created as an area becomes urbanized.
Percent of impervious surface has been recognized as a key indicator of impacts to watersheds due to
urbanization. Arnold & Gibbon1 (1996) discussed the two major advantages of impervious surface as an
environmental indicator. First, impervious surface is measurable and readily estimated. Second, impervious surface
is a major component of pollutant-generating land uses as well as a principal contributor to hydrologic change in a
watershed. According to Schueler2 (1991), the percentage of impervious surface is an indicator of stream channel
stability and stream health, and correlates highly with urban watershed pollutant loads. By determining the current
impervious surface cover and projecting future impervious cover potential, it is possible to gain a more clear
understanding of the potential impacts of this aspect of urbanization on the aquatic resources in the study area.
The Clackamas County Planning Department developed a series of land use design types or land use categories
as part of the Damascus-Boring Concept Plan development process. These land use design types are composed of
land use type polygons used in differing patterns across the landscape to provide development alternatives for the
Damascus-Boring planning area.
In order to estimate the amount of impervious surface that could be created with each development alternative,
it was necessary to calculate the percent of impervious surface of similar land use types in the existing urbanized
portion of the County. A series of thirty-eight land use polygons were identified using the orthophotography
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
developed from the LiDAR Digital Surface Model (DSM). These land use polygons were categorized based on their
representative land use design type in the Damascus-Boring Concept planning process.
IMPERVIOUS SURFACE DELINEATION
Project Requirements
Using the existing LiDAR intensity data and ½’ pixel orthophotography, WES contracted with i-TEN
Associates, Inc., Portland, OR, to create a detailed GIS-based impervious surface layer. Impervious surface
polygons were mapped to the tax-lot level for the entire Clackamas County Service District and the Damascus UGB
expansion area. Each impervious surface polygon was categorized by type of impervious surface, such as: Building,
Paved Road, Paved Drive, Paved Parking, Curb, Sidewalk, Slab, Pool, Bike Path, Athletic Field, Steps, and
Miscellaneous.
The contract specifications required the following deliverables:
ƒ 3D Shape files of Impervious surfaces
ƒ .E00 coverages of Impervious surfaces
ƒ Volume of Tree Canopy
2D Compilation
All the visible impervious surfaces were compiled in a MicroStation environment using a “heads-up”
digitization approach from the half-foot pixel color orthophotos (Figure 2). Each feature was captured in 2D using
a pre-defined and client approved data model for collection and data handling. The data collection process paid
particular attention to feature coding and coincident geometry of shared line features. Once the data collection was
complete, the data was programmatically isolated per feature type and translated. Using a largely programmatic
process, the data was migrated from the collection environment to the final data model – including attributation and
topologic building. Finally, the data was visually verified by skilled mapping technicians and QC specialists for
attribute and logical completeness.
Figure 2. Impervious Surfaces on Orthophoto.
Figure 3. Orthophoto from Intensity Data with
Impervious Surfaces overlay.
Nearly all digital orthophotography contains some level of radiometric displacement. In locations where the
radial displacement prevented accurate capture of impervious surfaces hidden by man-made strucutures or natural
features, the LiDAR intensity data was used to supplement the digital orthophotography (Figure 3). LiDAR
intensity data is orthographic in nature, therefore radial displacement is not present; it also has the same spatial
reference since the LiDAR surface data was used in the digital orthophoto production process. Utilization of the
intensity data allows an accurate delineation of features that were previously hidden in the orthophoto.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Creating 3D Shape Files
The impervious features captured in MicroStation from the orthophotos were converted to coverages using
ArcInfo . These coverages were then converted to Shape files using ArcView . (Figure 4)
A TIN was created from the “bare-earth” ASCII LiDAR data set using 3D Spatial Analyst (Figure 5). Once
the TIN was created, draping the 2D shape file onto the TIN created 3D shape files (Figure 6).
Figure 4. Shape Files Overlaid on Orthophoto.
Figure 6. Shape Files Draped onto LiDAR TIN.
Figure 5. TIN created from Bare Earth
LiDAR Data.
Figure 7. Building Polygons with
Assigned Elevations.
Building Elevations
Draping the building layer onto the “Noise” of the LiDAR dataset, then attributing the minimum elevation of
the points falling within each polygon derived the heights of the buildings. The non-ground LiDAR points were
converted to shape files with the Z value as an attribute.
Using these shape files, the LiDAR points were then linked to the buildings using the spatial join processes in
ArcView . The minimum Z value of the LiDAR points inside each building polygon was determined (Figure 7),
and the elevation values of the LiDAR points were assigned as an attribute to the building polygons. The 3D
building shape files were then merged into the planimetric 3D shape file.
Canopy Volumes for Wooded Areas
Using the orthophotos, closed polygons were digitized around the apparent edge of the tree canopies (Figure
8). From the raw LiDAR data, two surfaces were created: the bare earth surface and the “first return” or “top of
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
canopy”. The “wooded area” polygons were then used as a fence to compute the volume between the two surfaces
using Terramodeler . These values were then placed as text inside the polygon feature and then the polygons were
attributed.
Figure 8. Outline of Apparent Tree Canopy.
METHODOLOGY FOR ESTIMATING IMPERVIOUS SURFACE FOR
DEVELOPMENT SCENARIOS
Each of the thirty-eight current land use polygons identified by County staff were overlaid with the impervious
surface polygons to analyze the amount of impervious surface associated with each land use type. An EXCEL
Figure 9. Example of Land Use Outline.
Figure 10. Percentage of Impervious Surfaces.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Figure 11. Example of Land Use Outline.
Figure 12. Percentage of Impervious Surfaces.
spreadsheet was developed, which broke out the impervious surfaces by type for each land use polygon. Areas (in
square feet) were calculated for each impervious surface type associated with each land use polygon, as well as an
overall impervious surface percentage. In addition to impervious surface data, street width and total length of streets
were obtained. The number of dwelling units per acre was calculated using the WES customer database for the
purpose of comparing densities between the existing land use polygons and the land use type polygons developed
during the Damascus-Boring planning process. Figures 9 and 11 outline different land use types, while Figures 10
and 12 illustrate the percentage of impervious surfaces within those land use types.
Full Resource Protection
Karen Buehrig, Clackamas County Planning Department, supplied the four land use development scenarios
developed for the Damascus-Boring study area to WES. Each land use design type polygon used in the four land
use development scenarios was assigned an impervious surface percentage by comparing it to the representative
current land use types delineated by County staff. In cases where the current land use type impervious surface
percentage varied greatly between example polygons, an average impervious surface percentage was applied. Each
development scenario was then overlaid with the natural resource protection GIS layers the County has developed
to identify remaining buildable lands. The resource protections overlaid include: Title 3 riparian buffers, wetland
buffers, enhanced natural areas (for alternatives A&D), and Title 3 slope restrictions. Butte top areas were assumed
to maintain the same level of development that is currently in these areas.
Impervious surface for the urbanizing portion of each watershed was calculated using a weighted average
approach. This approach weights the amount of impervious surface generated by each land use type, by the
percentage of the total urbanizing area (buildable lands) in the basin that each land use type occupies. These
weighted percentages are then summed to create an overall impervious surface percentage for the urbanizing portion
of each watershed.
Percent Impervious for Urbanizing Lands
Table 1 illustrates the calculated impervious surface percentages of the urbanizing portion of one of the most
heavily urbanizing watersheds under development alternative E. The total impervious surface percentage is listed in
the last column. Only the urbanizing portion of the watershed was analyzed. The urbanized portion of the watershed
will be directly or in-directly connected to the stormwater network and thus will be source for the vast majority of
storm water runoff and pollutant loads to the receiving streams. Impervious surfaces associated with urbanized
landscapes is indicative of water quality in terms of both the rate of urban runoff and its pollutant constituents at the
“end of the pipe” and not necessarily at the bottom of the basin.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Table 1. Percent Impervious of Buildable Lands in the Sunshine Basin.
BASIN
SUNSHINE
URBANZING
ACRES
LAND USE
TYPE
ACRES
IMPERVIOUS
%
% OF URBAN
AREA
1145
RESB
UFF
ME
NC
PARK
RESA
RESA1
RESC
SCHOOL
93
509
190
25
75
86
118
35
12
1143
61
2
72
52
7
52
62
42
37
0.0312
0.4445
0.1659
0.0218
0.0655
0.0751
0.1031
0.0306
0.0105
0.9983
TOTALS
%
IMPERVIOUS /
% OF AREA
4.95
0.89
11.95
1.14
0.46
3.91
6.39
1.28
0.39
31.35
Preliminary Interpretation of Results
These results do not uncover a high degree of variability between development alternatives within an individual
basin in terms of their potential impact to aquatic resources. The Center for Watershed Protection has developed a
simple Impervious Cover Model (Figure 13). This model classifies watersheds into three categories, based on the
Figure 13. The Impervious Cover Model – Center for Watershed Protection.
percentage of impervious cover: sensitive, impacted, and non-supporting (Table 2). Much of the research that went
into the development of this model was performed in the Pacific Northwest and Mid-Atlantic ecoregions.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
Table 2: The Impervious Cover Model – Center for Watershed Protection
Subwatershed
Description
Category
Categories Derived from the Impervious Cover Model
Sensitive Stream
Subwatershed typically has impervious cover of zero to 10 percent.
Streams are of high quality, and are typified by stable channels, excellent
habitat structure, good to excellent water quality, and diverse communities
of both fish and aquatic insects. Since impervious cover is so low, they do
not experience frequent flooding and other hydrological changes that
accompany urbanization.
Impacted Stream
Subwatershed typically has impervious cover ranging from 11 to 25%,
and shows clear signs of degradation due to watershed urbanization.
Greater storm flows begin to alter the stream geometry. Both erosion and
channel widening are evident in alluvial streams. Stream banks become
unstable, and physical habitat in the stream declines noticeably. Stream
water quality shifts into the fair/good category during both storms and dry
weather periods. Stream biodiversity declines to fair levels, with the most
sensitive fish and aquatic insects disappearing from the stream.
Non-Supporting
Subwatershed impervious cover exceeds 25%. Streams in this
category essentially become a conduit for conveying storm water flows, and
Stream
can no longer support a diverse stream community. The stream channel is
often highly unstable, and stream reaches can experience severe widening,
down-cutting and streambank erosion. Pool and riffle structure needed to
sustain fish is diminished or eliminated, and the stream substrate can no
longer provide habitat for aquatic insects, or spawning areas for fish.
Water quality is consistently rated as fair to poor, and water contact
recreation is no longer possible due to the presence of high bacterial levels.
The biological quality of non-supporting streams is generally considered
poor, and is dominated by pollution tolerant insects and fish.
CONCLUSION
Using the Impervious Cover Model as a guide, all of the development alternatives result in each of the
watersheds being placed in either the impacted or non-supporting stream category with the exception of the
Richardson-Clackamas basin. It is important to note that the Impervious Cover Model is a predictive model and
does not take into account the affects of storm water treatment practices, riparian forest cover, and pollutants
generated from pervious surfaces.
Regardless of the development alternative chosen, the amount of impervious surface in each basin will increase
substantially with urbanization. None of the alternatives stand above the rest in terms of limiting the impact of
urbanization on the aquatic resources. At best, the alternatives redistribute the level of impact, with a few
watersheds consistently bearing the brunt of urbanization, most notably: Sunshine, Noyer, Richardson, and Rock
Creek basins.
Mitigation of the impact of these high impervious surface percentages on the aquatic resources will be difficult
at best and will pose an enormous challenge to storm water managers. Current technology and best management
practices will not be enough to fully mitigate the affects of development on water quality in the area. Efforts beyond
current management practices will be necessary to preserve water quality.
The solutions to these challenges will require a substantial departure from traditional public works and
development standards. New systems, techniques, and practices will need to be developed, most of which do not
exist today. The Pacific Northwest, and Oregon in particular, is on the leading edge of sustainable development,
using creative solutions for storm water management. Using advanced technology, such as LiDAR, we are hopeful
that with creativity and forward thinking we can find effective techniques and methods that will result in developing
sustainable urban communities that also supports healthy creeks and streams.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota
References
1
Arnold, C.L.; and Gibbons, C.J., (1996). Impervious surface coverage – The emergence of a key environmental
indicator. American Planners Association Journal , 62: 243-258.
2
Schueler, T.R. (1991). Mitigating the adverse impacts of urbanization on streams: A comprehensive strategy for
local governments. In: Proceedings of a National Workshop on the Integration of Stormwater into Local
Nonpoint
Source Issues, Northern Illinois Planning Commission, Chicago, Illinois.
Pecora 16 “Global Priorities in Land Remote Sensing”
October 23-27, 2005 * Sioux Falls, South Dakota