CHAPTER

10
Monitoring Applications

GIS is ideally suited to install, maintain, and query monitoring
equipment such as, rain gauges, flow meters, and water quality
samplers for system physical and hydraulic characterization. GIS
allows display and analysis of monitoring data simply by clicking on
a map of monitoring sites.

Gauging and inspection stations in the city of Los Angeles, California.

2097_C010.fm Page 177 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
LEARNING OBJECTIVE

The learning objective of this chapter is to familiarize ourselves with GIS applications
in monitoring data for effective operation and management of water, wastewater,
and stormwater systems.

MAJOR TOPICS

• Rainfall monitoring
• Satellite and radar rainfall data


NPDES

National Pollution Discharge Elimination System (U.S.)

NWS

National Weather Service

OGC

Open GIS Consortium

SCADA

Supervisory Control and Data Acquisition Systems

WSR

Weather Surveillance Radar

This book focuses on the four main applications of GIS, which are mapping, monitor-
ing, modeling, and maintenance and are referred to as the “4M applications.” In this chapter

we will learn about the applications of the second

M

(monitoring).


include closed-circuit television (CCTV) inspection of pipes, manhole inspections,
and smoke-testing of buildings. GIS applications for these types of data are described
in Chapter 15 (Maintenance Applications).
Hydraulic characterization data describe the quantity and quality of flow through
pipes and open channels as well as meteorological factors impacting the flow, such as
precipitation.
Wastewater and stormwater systems typically require data on flow quantity
(depth, velocity, flow rate, and volume), quality (e.g., suspended solids and bacteria),
and rainfall. Figure 10.1 shows a flowmeter and weir installation in a combined
sewer system overflow manhole. The flowmeter (Flo-Dar from Marsh-McBirney,
Inc.) shown on the left records incoming combined sewage depth, velocity, and flow
data. The weir shown on the right collects outgoing overflows. Some wastewater
and stormwater hydrologic and hydraulic (H&H) models require data on addi-
tional meteorological parameters such as ambient temperature, evaporation rate,
and wind speed. Water distribution systems typically require water pressure and
water quality data. In many projects, monitoring tasks make up a significant portion
of the scope of work and could cost 20 to 30% of the total budget. Careful installation
of monitors and effective management of monitoring data is, therefore, highly
desirable for the on-time and on-budget completion of monitoring projects.
GIS is ideally suited for selecting the best sites for installing various hydraulic
characterization monitors. Once the monitors have been installed, GIS can be used
to query the monitored data simply by clicking on a map of monitoring sites. GIS
can also be used to study the spatial trends in the monitored data. GIS is especially
useful in processing and integrating radar rainfall data with H&H models of sewage
collection systems and watersheds. This chapter will present the methods and exam-
ples of how to use GIS for installing and maintaining the monitors, and for querying
and analyzing the monitoring data.

REMOTELY SENSED RAINFALL DATA


(ASCE, 1999).

Figure 10.1

Flowmeter and weir installation in a manhole for monitoring incoming and out-
going flows.

2097_C010.fm Page 180 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
Radar Rainfall Data

Weather radars provide quantitative estimates of precipitation, which can be used
as input to H&H models. Radar rainfall estimates augmented with data from sparse
rain-gauge networks are useful in H&H modeling. Weather radars provide real time,
spatially distributed rainfall data that can be extremely valuable for flood forecasting
and flood warning.

NEXRAD Rainfall Data

The U.S. National Weather Service (NWS) has a group of weather radars called
the Next Generation Weather Radar (NEXRAD) system. NEXRAD comprises
approximately 160 Weather Surveillance Radar–1988 Doppler (WSR-88D) sites
throughout the U.S. and selected overseas locations. This system is a joint effort of
the U.S. Departments of Commerce (DOC), Defense (DOD), and Transportation
(DOT). The controlling agencies are the NWS, Air Weather Service (AWS), and
Federal Aviation Administration (FAA), respectively. Level II data provide three
meteorological base data quantities: reflectivity, mean radial velocity, and spectrum

Copyright © 2005 by Taylor & Francis
• Composite Reflectivity (CR): A display of maximum reflectivity for the total
volume within the range of the radar. This product is used to reveal the highest
reflectivities in all echoes, examine storm structure features, and determine the
intensity of storms.
• Echo Tops (ET): An image of the echo top heights color-coded in user-defined
increments. This product is used for a quick estimation of the most intense
convection and higher echo tops, as an aid in identification of storm structure
features, and for pilot briefing purposes.
• Hail Index Overlay (HI): A product designed to locate storms that have the
potential to produce hail. Hail potential is labeled as either probable (hollow green
triangle) or positive (filled green triangle). Probable means the storm is probably
producing hail and positive means the storm is producing hail.
• Mesocyclone Overlay (M): This product is designed to display information regard-
ing the existence and nature of rotations associated with thunderstorms. Numerical
output includes azimuth, range, and height of the mesocyclone.
• One-Hour Precipitation (OHP): A map of estimated 1-h precipitation accumu-
lation on a 1.1 nmi

×

1.1 nmi grid. This product is used to assess rainfall
intensities for flash flood warnings, urban flood statements, and special weather
statements.
• Severe Weather Probability Overlay (SWP): A measure of a storm’s relative
severity as compared with those around it. The values are directly related to the
horizontal extent of vertically integrated liquid (VIL) values greater than a

NEXRAD data and various visualization and analysis software tools are
available from NCDC and commercial vendors.

2097_C010.fm Page 182 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
Estimating Rainfall Using GIS

Because rainfall is a critical component in conducting H&H analyses, the
quality of rainfall data is critical for accurate system hydraulic characterization.
It is often the case that the spatial distribution of rain-gauges over a collection
system is too sparse to accurately estimate the rainfall over a given basin. Radar
rainfall technology can be used to obtain high-resolution rainfall data over an
approximately 2 km

×

2 km area (called a pixel). GIS can be used to display and
process the radar rainfall data.
NEXRAD Level III data from Doppler radar measurements provide spatially
dense rainfall data. These data are similar to those commonly seen on weather maps.
Such data do not require interpolation between point data from widely scattered rain-
gauges because they provide continuous rainfall measurements throughout a watershed
or sewershed (Slawecki et al., 2001).
WRS-88D radar images have a mean resolution of 4 km

×


WSI Corporation, a supplier of weather data in the U.S., provides the WSI
PRECIP
data over these pixels in 15-min time increments. These data can be obtained

2097_C010.fm Page 183 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
through commercial vendors and adjusted to reliable ground gauges, if necessary.
Hence, the radar rainfall volume adjusted through calibration with ground gauge,
combined with high-resolution spatial distribution, provides modelers with a very
accurate local estimate of rainfall. These estimates can be used with their flowme-
tering to characterize the basin of interest (Hamid and Nelson, 2001).

Radar Rainfall Application: Virtual Rain-Gauge Case Study

Precipitation varies with time and across a watershed. GIS can easily create a
rainfall contour map from point rainfall values. Unfortunately, point rainfall values
are generally scarce due to inadequate rain-gauge density. Many watersheds have
few or no rain gauges to accurately measure the spatial rainfall distribution. In this
case, a hydrologist may have to use the precipitation data from a nearby watershed.
Many watersheds do not have recording-type rain gauges to measure the hourly
variation of precipitation. In this case, a hydrologist may need to apply the temporal
distribution of a nearby watershed. The recent availability of radar precipitation data
via the Internet eliminates these problems by providing subhourly precipitation data
anywhere in the watershed.

×

1 km pixel of the rainfall atlas is equivalent to a virtual
rain-gauge. The users can click on any of the 2276 pixels to retrieve the 15-min
rainfall data in HTML or spreadsheet (Excel) format. This is equivalent to having
2276 virtual rain gauges in the County. This powerful tool revolutionizes collection
system planning, modeling, and management by providing the critical missing link
in urban hydrology — accurate local precipitation data. The rainfall atlas for a water-
shed near the city of Pittsburgh is shown in Figure 10.3. The figure also shows the
clickable pixel cells and retrieved 15-min-interval rainfall data for November 2000
for one pixel.
Vieux & Associates (Norman, Oklahoma) provides basin-averaged radar hyeto-
graphs at user-defined time increments (5, 15, 30, 60 min, or daily). Spatial resolu-
tions smaller than 1 km with data precision at 256 levels can be achieved. A real
time archive provides access to the radar rainfall estimates. Data are provided in

Figure 10.3

Radar-based virtual rain-gauge map and retrieved rainfall data.

2097_C010.fm Page 185 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
GIS-ready (Shapefile) and hydrologic-model-ready formats. Figure 10.4 shows a
sample image from the RainVieux Web site tailored for 3RWDDP. It provides historic
and real time radar rainfall measurements that can be queried from a Web site. Radar
and rain-gauge measurements are used for cost-effective design of sewer rehabili-
tation projects in the Pittsburgh Metro area.

The city of Huntington, WV, has a combined sewer system with 23 combined
sewer overflow (CSO) discharge points permitted under the NPDES program. The
sewer system is operated by the Huntington Sanitary Board (HSB). One of the CSO
requirements specified in the NPDES permit was to monitor each CSO event for cause,
frequency, duration, quantity, and quality of flow. As a first alternative to complying
with this requirement, the feasibility of monitoring all the 23 CSO locations was
studied in 1993. This option required purchase, installation, monitoring, and mainte-
nance of flow monitors, water quality samplers, and rain gauges for each CSO site.
This option was ruled out because of its excessive cost, estimated at over a million
dollars, and other problems such as adverse environmental conditions in the CSO
structures, difficult access issues, and new federal requirements for confined space
entry. The second option consisted of a combination of monitoring and modeling. In
this option, a representative subset of CSOs can be monitored temporarily to collect
sufficient calibration data and develop a calibrated model for each monitored CSO
area. Calibrated model parameters could subsequently be applied to the models of
unmonitored CSO areas. CSO models can eventually be used to predict the quantity
and quality of CSO discharges from observed rainfall data. This option, costing
approximately one third of the first option, was preferred by both the HSB and the
EPA (Region V), and was selected for implementation.
In 1994, ArcView was used to select a representative subset of the 23 CSO sites
to be monitored. Selecting the representative sites is a complex process. The objective
is to maximize the extent and value of the data for use in model calibration and
provide a uniform distribution of flowmeters, samplers, and rain gauges throughout
the system so that all parts of the service area are equally covered. Other factors
such as providing coverage to all land uses, known problem CSO areas, and all types
of diversion chambers, as well as access for installation and maintenance of moni-
toring equipment should be thoroughly studied. Using these selection criteria, a total
of six sites were finally selected for sampling and monitoring. Figure 10.5 presents
an ArcView plot showing the monitored vs. unmonitored CSOs and CSO areas. The
map indicates that by monitoring only 25% of the total number of CSOs, approxi-

range of SCADA systems with Haestad Methods’ WaterCAD, WaterGEMS, and Sewer-
CAD software programs. Using SCADA Connect, information can be automatically
queried from the SCADA system and passed to the WaterGEMS database, eliminating
manual processes. For example, tank levels and pump and valve settings can be updated
in WaterGEMS based on the SCADA database, so that the next time WaterGEMS is
run, it will reflect the current field conditions. SCADA Connect has been deployed for
the City of Bethlehem, Pennsylvania’s Bristol Babcock SCADA system and City of
Casselberry, Florida’s Citect SCADA system (Haestad Methods, 2004).

NPDES-PERMIT REPORTING APPLICATIONS

The well-known GIS magazine

Geospatial Solutions

holds an “Applications
Contest” each year (Geospatial Solutions, 2001a). The entries are judged based on

Figure 10.5

Flow-monitoring sites selected by using GIS.
2
3
4
22
23
24
5
6
7

9
10
11
12
13
14
15
16
18
19
20
21
17

Legend2
3
Monitored CSO Areas
Monitored CSOs (Flow
Meters and Samplers)
Unmonitored CSOs
Rain Gages
Treatment Plant
Unmonitored CSO Areas
Receiving Water

2097_C010.fm Page 188 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis

identify the inches of rainfall at each monitoring station. The application then
generates a rainfall grid using the tabular data for the specified rainfall year and
event to calculate the pollutant load. The final parameter the user specifies is the
pollutants. A dialog box enables interactive selection of the 256 pollutants (the
program automatically selects a default set of 64 pollutants). Once all the parameters
are defined, the user can calculate the pollutant load and generate maps and reports
for print using built-in templates. The program also automatically saves the selected
parameters and model results in a database table so that others can retrieve the load
run without having to reset all the parameters.
The project team used ESRI’s ArcView and Spatial Analyst for GIS analysis and
an automated report generation process from Crystal Decisions’ Crystal Reports.

MONITORING VIA INTERNET

Integration of GIS and Internet technologies is allowing users to click on a GIS
map and see the real time data from any type of Web-connected sensor, from stream
gauges to space and airborne Earth-imaging devices. Referred to as the Sensor Web,
the sharing of sensor data via the Internet represents a revolution in the discovery

2097_C010.fm Page 189 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
and use of live monitoring data. For example, data from the SAIC (a research and
engineering company in San Diego, California) Weather Monitoring System is
supplied from an Open GIS Consortium (OGC) Web Collection Service to a mapping
client, overlaying geographic information coming from an OGC Web Map Service.
The weather data are encoded in OGC Observation and Measurements Language,
which is an extension of Geography Markup Language (GML) (Reichardt, 2004).

in its distribution system, found that interpretations of water quality problems based
on aggregate monitoring data can be very misleading unless analysis is performed
at the appropriate scale. Incorporation of system and monitoring-site physical char-
acteristics data as well as the monitoring results into a GIS could have saved them
considerable investigatory effort (Schock and Clement, 1995).

USEFUL WEB SITES

Crystal Decisions www.crystaldecisions.net
Geospatial Solutions www.geospatial-online.com
Meteorlogix www.meteorlogix.com/GIS
NCDC Radar Products www.ncdc.noaa.gov/ol/radar/radarproducts.html
NEXRAIN Corporation www.nexrain.com
Vieux & Associates www.vieuxinc.com
WSI Corporation www.wsi.com

2097_C010.fm Page 190 Monday, December 6, 2004 6:06 PM
Copyright © 2005 by Taylor & Francis
CHAPTER SUMMARY

This chapter presented the methods and examples of GIS applications in install-
ing and maintaining various types of hydraulic characterization monitors (e.g., flow-
meters and rain gauges) and querying and analyzing the monitoring data. The
examples and case studies presented in this paper indicate that GIS is ideally suited
for selecting the best sites for installing the monitors. Once the monitors have been
installed, GIS allows user-friendly query of the monitoring data simply by clicking
on a map of monitoring sites. GIS facilitates studying the spatial trends in the