USING RASTER AND VECTOR GIS DATA FOR COMPREHENSIVESTORM WATER MANAGEMENT
Stephen P. Phipps(1)
ABSTRACT: Many communities are developing comprehensive storm water management
programs--funded by storm water utilities--to comply with Environmental Protection Agency (EPA)
regulations and the National Pollutant Discharge Elimination System (NPDES) permitting process. The
presenter will discuss how to use raster and vector GIS data to establish a storm water utility, manage
storm water infrastructure, and conduct basin master planning.
This paper will illustrate the many opportunities and benefits of using raster and vector GIS data for
comprehensive storm water management--including faster, more cost-effective decision making.
INTRODUCTION
Many communities are developing comprehensive storm water management programs--funded by
storm water utilities--to comply with Environmental Protection Agency (EPA) regulations and the
National Pollutant Discharge Elimination System (NPDES) permitting process.
Developing a comprehensive storm water management program requires three primary components,
among others:
Establishing a Utility
A continuing revenue source for storm water management can be established with a storm water utility,
based on service fees or user charges. Local governments that have established utilities have found the
method to be effective, reliable, and equitable because residents, businesses, and industries pay based
on their contribution to runoff and receive benefits from corresponding improvement projects.
While flat rates, based on equivalent residential units (ERUs), are typically assigned for most residential
parcels, utility service charges for nonresidential parcels are frequently based on impervious-area
calculations, often made using data from aerial photographs.
Creating an O&M Program (Infrastructure Management)
Ongoing O&M is also critical to successful storm water management. If not maintained, even a
well-designed and constructed storm water system will not perform as intended. Proper maintenance
includes periodic cleaning of inlets and basins and mowing and repair of land covers. However,
because of the age of storm structures in many cities and the lack of detailed engineering data and
location maps, public works officials often find that the location and condition of many storm water
structures is a mystery to them and to private property owners. Thus, the first step in implementing an
O&M program is establishing the current locations and conditions of storm water structures in the
community.
Conducting Modeling and Master Planning
A master drainage plan, coupled with computer modeling, is an important element in a comprehensive
storm water program. The master drainage plan should address short- and long-term drainage system
improvements covering basins within the incorporation and other basins that drain into the local
watershed. Developing the modeling and the associated plan requires an up-to-date, accurate, and
efficient mapping system.
Using Raster and Vector GIS Data
Communities pursuing a comprehensive storm water management program have found that raster and
vector GIS data can be combined efficiently and cost-effectively to help establish a storm water utility,
manage storm water infrastructure, and conduct modeling and basin master planning.
By creating a storm water GIS with both raster and vector data, communities have discovered that
implementing a comprehensive storm water management program and complying with EPA/NPDES
regulations is not as painful as they once thought.
USING RASTER AND VECTOR GIS DATA FOR UTILITY SETUP
Once a community has received an NPDES permit or has committed to establishing a comprehensive
storm water management program, a utility is often created to begin funding the storm water plans and
programs. Raster and vector GIS data can be used not only to calculate impervious areas and establish
utility billing rates for nonresidential parcels but also to identify utility billing credits for storm water "best
management practices" (BMPs).
Identifying Impervious Areas
1. Using Airborne Digital Imagery and Spectral Analysis to Identify Impervious Areas
Airborne digital imagery can be used economically and effectively to identify impervious areas. Red,
green, and near-infrared color bands--which include both true-color and color-infrared data
sets--would have to be collected to distinguish impervious areas from pervious areas; impervious areas,
for example, would appear in red on a workstation. Imagery can be captured at a resolution of 0.5 feet
per pixel. While such files are large, they are still manageable using today's storage technologies such as
RAIDS (Redundant Array of Inexpensive Disks).
ERDAS image processing/spectral analysis software is often used to automatically generate polygons
(raster data) of impervious areas, which can then be translated to ARC/INFO for further manipulation.
The next step: digitizing parcel lines (vector data) based on maps and other data usually collected from
the local auditor or assessor's office. Parcel lines input into ARC/INFO can then be overlaid with the
impervious-area polygons; the square footage of impervious areas in each nonresidential parcel can
then be calculated and passed on to the storm water utility billing system so that appropriate fees for
commercial and industrial parcels can be assessed.
Using raster and vector data to calculate impervious areas in this way is efficient because polygon
processing is automatic; heads-up digitizing of impervious-area polygons is not required. This not only
saves time but potentially produces fewer input errors.
2. Using Digital Orthophotos to Identify Impervious Areas
Digital orthophotos, raster images of ground features in their true map positions, are another alternative
for identifying impervious areas of nonresidential parcels. Once impervious areas are identified, the
digital orthophotos can be used as a strong foundation or backdrop for establishing a storm water GIS,
which can be used for storm water system O&M and as well as modeling and basin master planning.
Digital orthophotos are created by removing image displacements on aerial photography through
computer image processing. The first step is acquiring aerial photography and obtaining sufficient
horizontal and vertical ground control to orient the photographs to known coordinates and ground
features. Fully analytic triangulation is performed to mathematically densify the ground control.
Aerial photographs are then scanned with a micron-based film scanner to produce continuous-tone,
raster images of the photography. (The emergence of digital cameras is slowly eliminating the need for
image scanning.)
The aerial photographs are used again to compile digital terrain models (DTMs), a series of X, Y, Z
coordinates that accurately depicts ground elevations. The DTMs must be both accurate and dense
enough to adequately define the terrain. Two kinds of points are input to create DTMs: breaklines,
which indicate an abrupt change in elevation, and spot elevations in a grid-type pattern (mass points).
The final step in digital orthophoto production is image rectification. The raster images of the
photography are overlaid with the DTMs and corrected based on ground coordinates so that image
displacements can be removed. The result: Accurate, rectified raster images of the aerial photography,
which allow users to work with "soft copies"--computerized maps.
The digital orthophotos, with impervious areas clearly visible, are then used to perform heads-up
digitizing of impervious areas as polygons. Parcel lines (vector data) from the local auditor or assessor's
office can be input into ARC/INFO as an overlay to the impervious-area polygons digitized from the
digital orthophotos.
The square footage of impervious areas in each nonresidential parcel can then be calculated and passed
on to the storm water utility billing system so that appropriate fees for commercial and industrial parcels
can be assessed.
Using Digital Orthophotos to Identify Utility Billing Credits
for Storm Water BMPs
Some communities have chosen to further enhance the equitability of their utilities by rewarding BMPs;
these communities issue storm water utility billing credits to parcels that have installed structures or
implemented programs that help alleviate drainage problems.
The following are three common practices that warrant storm water utility credits in some communities
with utilities:
1. Practices That Reduce Discharge Peak. A credit may be given for structures that decrease or
slow down storm water from gaining access to the city's storm water system. Ponds or detention
structures are examples of such structures. Credits are awarded based on the surface square footage of
the pond or detention structure (as calculated from digital orthophotos, for example) and the amount of
water trapped as determined by engineering calculations.
2. Practices That Reduce the Discharged Volume to the Surface Drainage System. A credit
may be given for structures such as French drains that use infiltration techniques to allow storm water to
reach the ground water system rather than the surface drainage/utility management system. Credits are
awarded based on the existence of such structures, which can be identified during field data collection.
3. Practices That Increase Water Quality Using BMPs. A credit may be given for structures that
allow a storm's initial runoff, called the wash-off load, to be stored long enough so that heavy metals
and other pollutants that affect water quality can be eliminated before they reach the municipality's
system. A wet detention structure is one such structure. Typically, the first half-inch of a storm event
must be detained for at least two hours for a credit to be allowed. Credits are awarded based on the
surface square footage of the wet detention structure (as calculated from digital orthophotos, for
example). Credits for BMPs can also be awarded by using raster data such as digital orthophotos to
calculate surface square footage of other structures that help alleviate drainage problems. Such
structures include constructed wetlands, Austin first-flush filtration basins, and grassed swales.
Raster and vector data used during utility setup also provides a side benefit for a community: an instant
customer-service system and a permanent record, based on impervious-area calculations, of how the
storm water bill for each property was derived. Customers who contact the municipality with questions
about their storm water bills will be able to see--via a map/photo on the screen or a plot of their
property--the basis for the impervious-area measurement and the storm water utility charge.
USING RASTER AND VECTOR GIS DATA FOR
INFRASTRUCTURE MANAGEMENT
Once a utility is established and a revenue stream begins, a community can start implementing an
ongoing O&M program for better managing its storm water infrastructure. Raster and vector GIS data
can be used to establish a spatial inventory of storm water features and link inventory features to a
third-party work-order processing system. These activities will help the community improve routine
maintenance and more easily address customer complaints and emergencies such as flooding.
A flooding or drainage problem can occur for several reasons:
Controlling flooding problems and initiating routine O&M means knowing the locations, conditions, and
other characteristics of structures within the community's storm water system. The first step: establishing
a strong GIS foundation for adding storm water system features, structures, and attributes. (If digital
orthophotos were used during utility development, they can also be used as the basis for adding storm
water graphic and nongraphic data. Otherwise, a base mapping foundation would have to be
established.)
Using Raster and Vector GIS Data to Establish a
Spatial Storm Water Inventory
If digital orthophotos have been created, a community can effectively combine both raster and vector
storm water data to establish a spatial inventory of storm water features and structures.
1. Creating Raster Data of Storm Water System Details
If fairly accurate and up-to-date storm water maps exist depicting fine details about the storm water
system (for example, exact manhole and inlet locations), they can be scanned and incorporated into the
GIS system. Scanning instead of digitizing/vectorizing allows more data to be converted in less time;
moreover, valuable storm water data that otherwise may not have been included will be available in the
storm water GIS for O&M applications.
2. Creating Vector Data of Major Storm Water Structures
Next, a basic GIS map of major storm water structures and features would be digitized as vector data.
(Digital orthophotos would be used as the accurate backdrop for adding storm water graphic features.)
The resulting map would include structures such as pipes, open channels, and ditches; ID numbers
would be assigned to all structures input.
The raster images containing fine details about the location of other storm structures would be available
as "visual attributes" to the major storm water structures and features. When detailed locational data
was needed on a particular storm structure, the raster image--linked to the structure via an ID
number--could be accessed via a point-and-click operation.
Using Field Methods to Establish a Spatial Inventoryof Storm Water Structures
Instead of converting existing storm water maps to raster and vector data, a spatial inventory of storm
water structures could also be established using field methods in conjunction with the accurate digital
orthophoto base map. This alternative would result in more detailed, current information than would be
available if only existing maps were used.
There are two options within the field-methods alternative:
1. The Total GPS Inventory Option
This option is preferred when existing storm water system maps are outdated. Inventory crews would
use GPS methods and hand-held data collectors to occupy storm water structures, determine
locations, and collect attributes needed for O&M, modeling, and master planning applications.
Structures to be located include manholes, pipes (inlets and outfalls), and culverts. Channels, including
streams and open channels, would also be located.
The result would be the locations (x,y coordinates) and attributes needed for each structure (such as
the elevation). Data collected in the field could then be downloaded, translated into ARC/INFO, and
overlaid and linked with the digital orthophoto base map. The ID number for each storm water
structure would be its x,y coordinate.
2. The Data Prep/Field Method Inventory Option
This option is preferred when existing storm water system maps are fairly accurate and current and tied
to an accurate base map (but not necessarily digital orthophotos). Storm water structures on the
existing hard-copy maps would be heads-up digitized into the ARC/INFO GIS; ID numbers would be
assigned automatically to storm water structures input. The ARC/INFO system would then be used to
plot a map containing the storm water structures and the newly assigned ID numbers.
Inventory crews would use these maps in the field to collect attributes on the identified storm water
structures via hand-held data recorders. Attributes collected and tied to the structure ID number on the
map would later be downloaded into the ARC/INFO GIS and linked to the graphic structure via the
ID number.
Even if storm water maps had been updated regularly, it is possible that inventory crews may find
structures not identified on the newly plotted maps. Therefore, such structures would be sketched on
the maps and assigned an ID number in the field based on a predetermined formula so that attributes
could still be collected and attached.
Using Raster and Vector GIS Data to Link Inventory Features
to a Third-Party Work-Order Processing System
Once a storm water inventory/GIS system is established, raster and vector data can also be used with
a work-order processing system to ensure that storm water complaints from customers are addressed
and resolved quickly.
A storm water structure's ID number provides the link to the work-order processing system. For
example, a dispatcher receiving a customer compliant via telephone would input the customer's address
into the GIS system; the address would be linked to its appropriate pipe ID numbers, which would
automatically tell work crews where they might begin their investigation of the drainage problem. If
desired, crews could also choose to plot a vector map of the affected area as well as storm water
detail maps on structures in the immediate area (if available in the GIS as raster images). But raster and
vector storm water GIS data can also be used for scheduling routine maintenance. Structure IDs within
a certain range could be selected and maps containing those structures plotted. O&M crews could
take these maps into the field to locate, clean, and perform any needed maintenance on targeted
structures. This method provides a consistent means to ensure all structures are properly and routinely
maintained.
In the past, a community may not have attempted a regular maintenance program for all structures
because of a lack of funding. This strategy for routine maintenance would allow communities--many for
the first time--to begin a proactive storm water management program. Such a program would not only
allow problems to be addressed as they occur but also prevent certain new drainage problems from
occurring at all. As a result, the number of flooding complaints and emergencies reported by customers
during a storm will likely decrease significantly.
USING RASTER AND VECTOR GIS DATA FOR
BASIN MASTER PLANNING
The raster and vector storm water GIS data collected for utility development and infrastructure
management can also be used for modeling and basin master planning: a solid foundation has been
established; storm water structures have been located and tagged with ID numbers; and attributes have
been collected. But before modeling can be performed and a master plan developed, DTMs must be
collected and included in the GIS.
Collecting DTMs
DTMs, a series of X, Y, Z coordinates that accurately depicts ground elevations, are required for
storm water modeling so that accurate, hydrologically correct surface models can be generated. (A
surface model is a digital 3-D representation of the ground surface.)
Having a hydrologically correct surface model lets the ARC/INFO system automatically cut channel
cross sections by "feeling" the digital topography. The resulting channel profiles show graphically how
much storm water the channels can handle before flooding occurs.
Aerial photography is used as the basis for generating DTMs. If digital orthophotos were produced as
the accurate backdrop for the storm water GIS, DTMs were created; however, the process used to
create DTMs for digital orthophotos and the process used to create DTMs for modeling applications is
not the same. DTMs used for digital orthophoto creation are typically not suitable for modeling
applications; however, if the digital orthophoto project was planned with the idea that DTMs would
eventually be used for modeling applications, it is possible that existing DTMs could be used.
If no DTMs are available, USGS digital elevation models (DEMs) could be used. However, USGS
DEMs have an accuracy of +/-30 meters; therefore, the resulting modeling data would only be reliable
for regional planning and not detailed basin master planning.
Linking ARC/INFO GIS Data with Hydrologic and Hydraulic Models Storm water GIS data already
collected could be used to perform hydrologic and hydraulic modeling. By linking the ARC/INFO
storm water GIS with HEC-1 (hydrologic) and HEC-2 (hydraulic) models, enhanced modeling
capabilities are available. Moreover, modeling becomes easier because manual calculations are not
required. For example, if the models were not linked to the GIS, manual calculations would be required
for the cutting of cross sections, which would take at least five times longer than using the ARC/INFO
system linkage.
ARC/INFO Data Required for Linkage with the HEC-1 and HEC-2Models
The following storm water structures should be collected in the field and included in the ARC/INFO GIS for modeling and master planning as well as the O&M application (see Table 1):
| Table 1. Storm Water Structures to be Collected in the Field for O&M, Modeling, and Master Planning | ||
| Structure | Example Attributes | |
| Streams/open
channels (and Ids) (for generating centerline data) |
| |
| Manholes (and IDs) |
|
|
| Pipes--inlets and outfalls--(and IDs) |
|
|
| Culverts (and IDs) |
| |
The following storm water features, collected or compiled from aerial photography sources and other
data, should be included in the ARC/INFO GIS for modeling and master planning as well as the O&M
application (see Table 2):
| Table 2. Storm Water Features to be Collected or Compiled from Aerial Photography Sources and Other Data for O&M, Modeling and Master Planning | ||
| Feature | Example Attributes | |
| Street centerlines (and IDs) (Intelligent centerlines can be used | Address | |
| Polygons of impervious areas (and IDs) (Only large paved contributors, | Square footage of impervious areas | |
| Digital topographic data (DTMs) | Not applicable | |
Characteristics of the Interface between ARC/INFO
and the HEC-1 and HEC-2 models
Woolpert has successfully linked the HEC-1 and HEC-2 models with ARC/INFO and implemented
the system's capabilities on several modeling projects. Primary characteristics of the interface include
the following:
1. The HEC-1 model uses vector data--streams and open channels--and the ARC/INFO
networking function to show the locations of and relationships between the different types of
flows: sheet flow, shallow channel flow, and concentrated channel flow. The linkage allows users
to see graphically, on a map, how fast the three flow types are moving at different locations within the
model (volume/time). These unit hydrographs, which make up the HEC output card, are then used to
feed the HEC-2 hydraulic model.
2. ARC/INFO can be used to graphically show the output of the HEC-2 model--the extent of
flooding for various storm events. A calibrated and verified HEC-2 model can also be used to
develop two-, 10-, 25-, 50-, 100-, and 500-year frequency flood profiles for streams and open
channels within the watershed. The linkage with ARC/INFO allows users to see these results
graphically.
3. The interface also allows for easier model calibration. Because the models are linked to
ARC/INFO, model calibration is easier because results are posted graphically and any needed changes
to cross sections can be made via a user interface. Calibration ensures that scenarios the model will
predict are based on results from actual rain events. To ensure model predictions will be realistic,
adjustments are required to correct cross sections, unit hydrographs, and other data.
Modeling Results
Once the modeling is performed, modeling results are typically summarized in a report. This report,
which could then be used as a starting point for basin master planning, might include the following:
Because the models were linked to ARC/INFO, channel cross-section locations and the two-, 10-,
25-, and 100-year frequency floodplains and flooded structures--delineated in different and
corresponding colors--could be plotted on maps along with base mapping data.
Once the modeling is performed and conclusions reached, the basin master plan can be developed. A
comprehensive basin master plan might include the following:
The HEC-1 and HEC-2 models, linked with ARC/INFO, can also be used to do "what if" scenarios
for basin master planning applications:
By modeling these scenarios, users will be able to determine what impact--positive or negative--a
proposed development will have on downstream structures.
CONCLUSION
Establishing a storm water utility, managing storm water infrastructure, and conducting basin master
planning to comply with EPA/NPDES regulations and meet the community's needs can best be
accomplished by using raster and vector GIS storm water data as an information management and
planning tool. Once a storm water GIS is established, more data can be added in the future to further
enhance the system's usefulness.
However, if a community chooses not to establish a GIS as it goes about implementing a
comprehensive storm water management program, all subsequent tasks will be significantly more
difficult. But more importantly: without a GIS to help organize, store, analyze, update, and manage such
large amounts of data, it's unlikely that communities will be able to successfully implement critical storm
water management plans and programs needed to properly manage the quality and quantity of storm
water.
ACKNOWLEDGEMENTS
The author wishes to thank Teresa Zumwald of Zumwald & Company in Englewood, Ohio, for her
assistance in writing and editing this paper.
REFERENCES
Brinkman, Robert F., C.P., Woolpert, Dayton, Ohio. "Digital orthophotos for utility applications,"
1992.
Reese, Andrew J., P.E., Ogden Environmental and Energy Services, Inc., Nashville, Tenn.
"Understanding storm water problems," in APWA Reporter, January 1993, pp. 7-8, 22.
Woolpert, Louisville, Ky. "Storm water management in the '90s," 1992.
1. GIS Project Manager, Woolpert, 409 East Monument Avenue, Dayton, Ohio, 45402-1261.