Joliet AAP Sewer Study – Group 3
Mobile GIS Application and Mapping
Overview
Joliet Army Ammunition Plant was
constructed during World War II for the purpose of
manufacturing, loading, assembling, packing, and shipping of
bombs, projectiles, fuses, and supplementary charges.
Although the plant was used extensively during World War II,
in 1945 all production of explosives was halted, the
sulfuric acid and ammonium nitrate plants were leased out;
and the remaining production facilities were put in layaway
status. The installation was reactivated during the Korean
War, and again during the Vietnam War. Production at the
plant gradually decreased until it was stopped completely in
1977. In April 1993, the property was declared as excess by
the Army and the land is being transferred to various
Federal, local and state jurisdictions.
A safety assessment of the land was
required which included a survey of manholes. ike was
identified as the ideal data capture device.
Application Development
A series of GIS forms were developed for
ESRI ArcPad using ArcPad Application Builder software
(version 6.0.1 for NT). The over-arching objective in using
mobile GIS is to have all pertinent data pre-loaded on the
handheld and all forms, prompts, dropdown lists, and
in-field references available. The forms were designed with
this in mind and included standardized, validated lists of:
unit of measure (SDSFIE), ground cover (custom list), and
direction (16 bearing selections based off cardinal
directions, e.g., N, N/NE, NE, E/NE, E, etc.).
Based on requirements, forms were built
for the following types of features known or anticipated to
occur at the Joliet site:
- 1. Building Footprint
- 2. Fence Line
- 3. Road
- 4. Ditch/swale
- 5. Storm Manhole
- 6. Sanitary Manhole
- 7. Storm Network Outfall
- 8. Building Drain
- 9. Building Trench
- 10. Building Pit
- 11. Building Sump
The majority of forms (1, 2, 3, 4, 7, 8,
9, 10, 11) were built off of a standard form that included
GPS received coordinates in UTM format, date-time stamp,
label, comment, and associated pictures taken. The Building
Footprint form, for example (Figure 3), contained these
attributes and allowed for five separate photos to be
associated with one point recorded. Forms were designed to
identify each point not only by UTM coordinates and
time-stamp but also by a unique ID number. Since the
Building Number was an assigned, historical reference to the
structures, it was later used for the label.
ArcPad Form for Building Footprint.
The manhole forms were the most complex
of the types of data required for collection. As such, each
of the sanitary and the storm manhole forms needed to have a
Summary and a Details page in addition to Image pages for
photographic recordation (Figure 4).
 
ArcPad Form for Sanitary Manhole, Summary & Details Pages.
Storm manhole forms were modeled in the
same fashion but simply pointed to a different set of SHP
files. The effort of the requirements meeting was to
anticipate as many types of data as possible, but allowance
had to be made for the unexpected; consequently, a Generic
form was added to the suite of forms so that basically any
land feature could be recorded and given a label,
photographed, and described in notation. If quantities of a
particular feature type that had not been given its own form
were encountered, then the Generic features could be parsed
to make those cases their own layer. This was later the case
for certain water network features, including hydrants and
valve controls.
Once all of the forms had been built,
effort was made to collect coverages for the basemap at
Joliet. USGS topographic maps were found and downloaded as
well as some aerial photographs, but there
were issues with projecting some of these. In the end, the
most useful basemap feature used in helping to identify
features on the landscape was a 1941 map of the proposed
sanitary and storm lines.
1941 Map of Proposed Sanitary and Storm
Networks.
While this map was not comprehensive
(i.e., it was missing an entire sanitary line on the
northwestern side of Group 3), it was fairly accurate on
many of the manholes in reference to the buildings shown.
Georeferencing the schematic onto the ikeTM
proved fruitless due to projection issues; the ikeTM
currently only operates within the WGS84 projection which is
the native language of the satellites’ GPS signal. This
issue was bypassed via ArcMap, whereby the image was
georeferenced in the NAD27 projection, proposed locations of
sewer manholes were mapped on their own layer over this map,
and then that layer was converted to the WGS84 projection
for the ikeTM. Hence, the green points as shown in
Figure 6 were taken into the field.
Proposed Manhole
Layer Over the1941 Plan on ikeTM.
Fieldwork with ike TM
The fieldwork for this project was
performed during at least four separate visits to the Joliet
site, and ikeTM was employed for data capture during
the first, third, and fourth visits. The first visit was the
initial data capture survey that lasted one week and
resulted in mapping of most buildings, roads, culverts,
manholes, and sewer networks (Figure 7).
Subsequent visits to the site augmented
this data set, at times providing only minor revisions to
the map, but on the third field visit an entire sanitary
line was identified. The fourth visit was for ground
truthing the final map.
The following maintenance routine
was employed daily during the field data capture portion of
the Joliet survey:
- 1. Charge ikeTM overnight.
- 2. Ensure that ikeTM has all previously
recorded GIS data loaded.
- 3. Capture field data, SHP file features and JPG
photographs.
- 4. Charge ikeTM during lunch.
- 5. Capture field data, SHP file features and JPG
photographs.
- 6. Download all SHP and JPG files from 2 ikeTM
units onto a laptop.
- 7. Convert projections to NAD27 from WGS84.
- 8. Consolidate the 2 data sets into one for each of
the layers.
- 9. Convert projections back to WGS84 from NAD27.
- 10. Upload all consolidated SHP files to the 2 ikeTM
units from the laptop.
At the time of the initial fieldwork, the
ikeTM was only capable of recording single points in
the mobile GIS. This eliminated the possibility at that time
of recording line or polygon features. Consequently, single
point features (e.g., manholes, hydrants, valve controls,
building drains) were recorded rapidly in the field and
required no further manipulation, but multi-point features
(e.g., sewer lines, roads, fences, buildings) required a
series of independent points recorded and adequately labeled
in the field to facilitate subsequent joins on a polygon or
polyline feature.
The following data capture routine
was employed with ikeTM in order to record various
features as required on site:
- 1. The ikeTM unit was initialized by the Power
button (upper right).
- 2. The ArcPad was launched (right function button).
- 3. The AABDCA application was launched (left function
button) which initialized the GPS receiver, the laser
distance meter, the camera, the compass.
- 4. The user would first click ‘Settings’ to ensure
that the declination was set to northeastern Illinois, or
-0.06 degrees.
- 5. The user would select from a dropdown list the
layer intended for recording the target feature (if not in
the original 11 lists, then ‘Generic’)
- 6. The user would aim the ikeTM cross-hairs at
a target shown in the display screen.
- 7. The user would ensure that the PDOP reading was
sufficient. Usually, a PDOP no greater than 3 was
enforced, and the North American WAAS real-time
differential correction signal was employed wherever
possible (display shows ‘SBAS’ when engaged).
- 8. The user would click the ‘Capture’ button to take
the photograph and record the GPS coordinates.
- 9. The user would click ‘Finish’ if the photo shown
was sufficient.
- 10. The user would fill in any additional form data
required or desired.
- 11. The user would complete the point logging by
clicking the ‘X’ in the upper right-hand corner of the
screen.
Determining where on the target one was
to aim the camera and laser distance meter depended upon the
feature itself. Single point features such as manholes and
hydrants typically received a single shot in the middle of
the object itself. Roads, fences, and buildings were
different. Mapping of roads required points selected on the
centerline where there was transition from a straight line
to a curve. Unlike other features, mapping of the fences
required close proximity so that the laser did not
overshoot, or shoot through, the weave.
Mapping of buildings required a single
point at each vertex. Since the laser distance meter on the
ikeTM has
to have a flat surface to reflect off, the exact corner of
the building could not be recorded. Additionally, in many
cases, the vegetation was so thick around the base of the
buildings that these points had to be shot either through
the understory at the base or in the upper corner of the
structure near the cornice. Figures 8 and 9 illustrate this
on Building 3-7 where 4 vertices were captured to record a
4-sided building. Note that the vegetation at the northeast
corner of the building mandated a footprint vertex recorded
at the structure’s cornice.
Photograph of Building
3-7 (View Northwest).
Map of Building
Footprints Recorded.
In total, 791 points were captured during
the field data collection using ike TM,
with an average of 22 points captured per hour. This count
preceded subsequent mapping work with ArcMap at which time
some features may have been reclassified after further
inspection of context and comments recorded. Table 1 shows
the breakdown of the raw points by date visited.
Table 1. Raw Points Captured with ikeTM
by Date.
|
DATE |
HOURS |
IKE 1 |
IKE 2 |
TOTAL |
PACE |
|
9/20/2004 |
5.0 |
84 |
11 |
95 |
19 |
|
9/21/2004 |
7.0 |
156 |
7 |
163 |
23 |
|
9/22/2004 |
7.0 |
125 |
9 |
134 |
19 |
|
9/23/2004 |
7.0 |
211 |
13 |
224 |
32 |
|
10/19/2004 |
6.0 |
7 |
7 |
1 |
|
10/20/2004 |
0.5 |
3 |
3 |
6 |
|
11/9/2004 |
4.0 |
165 |
165 |
41 |
|
TOTAL |
36.5 |
751 |
40 |
791 |
22 |
The hours represented above indicate only
the periods during which ike TM
was in use. As described in the field methods for manhole
and sewer identification, there were periods during the
field research that required no immediate need for mobile
GIS, particularly when the smoker was in use or technicians
were in transit to another portion of the site.
The raw data can be further subdivided
into the categories, sanitary manholes, storm manholes,
buildings, generic features (hydrants, valve controls), and
other (building drains, pits, sumps, trenches, road
vertices, fence vertices) (Table 2).
Table 2. Raw Points Captured with ikeTM
by Category.
|
DATE |
SAN |
SW |
BLDG |
GEN |
OTHER |
TOTAL |
|
9/20/2004 |
3 |
2 |
27 |
8 |
55 |
95 |
|
9/21/2004 |
16 |
4 |
78 |
21 |
44 |
163 |
|
9/22/2004 |
1 |
16 |
52 |
25 |
40 |
134 |
|
9/23/2004 |
9 |
148 |
19 |
48 |
224 |
|
10/19/2004 |
3 |
4 |
7 |
|
10/20/2004 |
2 |
1 |
3 |
|
11/9/2004 |
151 |
14 |
165 |
|
TOTAL |
25 |
35 |
305 |
224 |
202 |
791 |
A total of 60 sewer manholes were
recorded during the project. At many of these manholes, the
ArcPad form was used to record attributes regarding the
manhole inlets and outlet. Bearings and diameters were taken
for each of the pipes connected to the manhole chamber;
these bearings were later used in ArcMap to connect the
manholes to either the sanitary network or storm network.
The 305 building vertices that were recorded were later used
to draw the polygonal building features.
By the time the crew returned to ground
truth the maps and data, the polyline and polygon mapping
capability had been added to the ike TM.
Figure 10 shows the ikeTM screen with the point,
line, and polygonal layers after the raw data had been used
to draw the actual sewer lines and the buildings.
Figure 10. Joliet Map of Point, Line, &
Polygon Layers on
ikeTM.
Mapping
ikeTM
Captured Data
After the raw data were collected from
the field, they were converted to the NAD27 projection. The
layers were then imported into ArcMap. There, the two
different ikeTM
data sets were consolidated into one for each of the layers
corresponding to the original list of required features (see
above, Information Requirement Item 2). Based upon these raw
point data, several different polyline and polygon layers
were created in order to accurately represent features at
the site. These included the following GIS layer types (with
SHP file name):
Polylines
- • Roads (roads)
- • Perimeter Fence (fence)
- • Sanitary Line (sanitary_line)
- • Sanitary Line Uncertain (sanitary_uncertain)
- • Storm Line (storm_line)
- • Storm Line Uncertain (storm_line)
- • Water Line Uncertain (water_line)
Polygons
- • Buildings (JolietBuilding)
- • Building Features (building_feature)
- • Sanitary Structures (sanitary_structure)
- • Storm Structures (storm_structure)
There were several iterations of map
creation and revision that occurred during the mapping stage
as errors in interpolation were identified and as additional
raw field data were collected. Interpolation was used
predominantly to draw the sanitary and storm networks and
was based upon the 1941 plan, raw point data captured with
the ike, and field notes.
Once the layers were defined and drawn,
they were grouped into functional categories: water system,
storm system, sanitary system, other fixed assets, and
sampling for explosives (Table 3). These layers were then
optionally made visible or invisible depending upon the map
that needed to be generated.
In total, the water system was comprised
of 13 manholes, 12 water hydrants, and 17 valve controls
interlaced in a network that encircled the group. The storm
system was comprised of 35 manholes, 23 related structures,
and an outfall set within a network that basically bisected
the group east-west and channeled storm discharge to both
the east and the north. The sanitary system was comprised of
25 manholes, 3 sanitary structures, and 7 septic tanks set
within a network of basically two lines feeding to the west.
The southern line flowed southeast-northwest, paralleling
the south perimeter road, and serviced buildings 3-20A, 3-8,
3-19, and 3-20B. The northwest line flowed west and
southwest from building 3-64A, past 3-66, to a connection
northwest of 3-3.
Table 3. Features Grouped by Category
|
Category : Feature |
Count |
Note |
Total |
|
Water System |
|
Water Manhole |
13 |
|
Hydrant |
12 |
|
Valve Control |
17 |
|
Water Line - Uncertain |
20 |
segments |
|
62 |
|
Storm System |
|
Storm Manhole |
35 |
|
Storm Structure |
23 |
|
Storm Outfall |
1 |
|
Storm Line |
55 |
segments |
|
Storm Line - Uncertain |
6 |
segments |
|
120 |
|
Sanitary System |
|
Sanitary Manhole |
25 |
|
Sanitary Structure |
3 |
|
Sanitary Septic |
7 |
|
Sanitary Line |
32 |
segments |
|
Sanitary Line - Uncertain
|
19 |
segments |
|
86 |
|
Other Fixed Assets |
|
Building |
89 |
|
Building Feature |
2 |
|
Monitoring Well |
4 |
|
Fence |
1 |
perimeter |
|
Road |
29 |
segments |
|
125 |
|
Explosive Sampling |
|
Sample Xwipe - Positive |
7 |
|
Sample Xwipe - Negative |
45 |
|
Sample Jar |
68 |
|
120 |
The resulting maps were generated using
ArcMap’s layout functionality, employing a title, the North
Arrow, a scale, and a legend. The maps were then printed on
a large scale plotter using the same software.
|
