Upload
others
View
5
Download
0
Embed Size (px)
Citation preview
Predictive Modeling of Overland Petrochemical Spill Trajectories Using
ArcGIS
by
Kristen M. Mathieu
A MAJOR PAPER SUBMITTED IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF MASTER OF ENVIRONMENTAL
SCIENCE AND MANAGEMENT
UNIVERSITY OF RHODE ISLAND
DECEMBER 9, 2011
MAJOR PAPER ADVISOR: Dr. Peter V. August
MESM TRACK: Remote Sensing & Spatial Analysis
2
Table of Contents
I. Introduction ……………………………………………………………………………………. 3
II. Goal of Project ………………………………………………………………………………… 5
III. Sources of Data ……………………………………………………………………………… 6
IV. Methods ………………………………………………………………………………………… 7
V. Results ……………………………………………………………………………………………. 22
VI. Discussion & Conclusion ……………………………………………………………….. 26
VII. Acknowledgements ……………………………………………………………………… 30
VIII. References .………………………………......…...…………………………………….. 31
3
I. Introduction
Geographic Information Systems (GIS) have long been used in conjunction with
remote sensing technology to map accidental petrochemical spills in the environment
(Li, et al 2000). In this capacity, GIS can be defined as “a geoinformation technology for
… storing and retrieving data and images, as well as for processing these data into
information for scientists, environmentalists, and decision‐makers (Ivanov &
Zatyagalova 2008).” From the creation of Environmental Sensitivity Indices to predictive
spill trajectory models, GIS has proven to be a powerful tool for predicting and tracking
the impacts of petrochemical spills (Sorenson 1995). One of the notable uses of GIS is
the ability to model and predict the trajectory that a liquid pollutant will take once it has
been released from a point source. While the majority of these models focus on the
trajectory of maritime oil spills (APASA 2003), it is also possible to determine the
trajectory pollutants will take on land through the use of GIS tools originally developed
for hydrologic modeling.
The purpose of this technical paper is to describe and discuss the methodology
used to create flow path maps for the purpose of predicting the trajectory of liquid
pollutants in the terrestrial environment. The first section of this paper will discuss the
impacts of hydrocarbons on the terrestrial environment and the need for accident
mitigation measures, while the second section of this paper will describe in detail the
steps necessary to create predictive flow path maps. Any individual with a background
in basic GIS procedures will be able to create their own flow path maps based on the
instructions provided in this section.
Petrochemical Spills in the Terrestrial Environment
Through numerous cases studies of maritime oil spills, the scientific community
has come to understand the wide range of short‐term and long‐term negative impacts
that petrochemical spills can have on the marine environment. While not as frequently
studied, the negative impacts of petrochemical spills in the freshwater and terrestrial
environments are also understood (Vandermuelen 1995). Diesel, gasoline, and motor oil
all contain heavy metals and polycyclic aromatic hydrocarbons, which are known
4
carcinogens and naturally long lived due to their high molecular weight. When these
pollutants are introduced to the natural environment, they tend to accumulate in
groundwater and sediments, where they neither dissolve nor evaporate (Lloyd &
Cackette 2001). Instead, they remain in the soil and the water, where they are toxic to
both flora and fauna. A case study in 1997 involving a massive diesel fuel spill near
Cayuga Lake in New York found that within 24 hours of the spill, 92% of the local fish
population was dead (Lytle & Peckarsky 2001). Additionally, sampling and observations
of the local benthic invertebrate community over the next fifteen months found that the
spill significantly reduced invertebrate density by 90% and overall species richness by
50%. By the end of the fifteen‐month period, species density had recovered, while
species richness had not (Lytle & Peckarsky 2001). A laboratory study investigating the
impacts of diesel fuel on select plant species found that generally, diesel fuel is
phytotoxic to plants even in low concentrations, and that plant germination and
development is negatively effected by exposure to soil that has been contaminated by
diesel fuel (Adam & Duncan 1999). Used motor oil has been found to be both mutagenic
and tetratogenic to American the green tree frog (Hyla cinerea), preventing the
metamorphosis of tadpoles and stunting tadpole growth (Mahaney 1994). Of course,
the cost of petrochemical spills isn’t just measured in terms of the plants and animals
are killed or sickened, or the acres of land and water that are irrevocably poisoned.
Cleanup efforts for large diesel and oil spills can last anywhere from years to decades,
and can cost millions of dollars to complete. In some cases, bioremediation can literally
take hundreds of years (Lloyd & Cackette 2001). With figures like those, there is little
wonder why government agencies and environmental groups alike work tirelessly to
mitigate potential petrochemical spills before they begin.
The Rhode Island Army National Guard
Because of the potentially devastating effects that a major gasoline, oil, or diesel
fuel spill could have on the environment, the Rhode Island Army National Guard is
required by both State and Federal law to have a Spill Prevention and Contingency Plan
for the Camp Fogarty Training Site, located in East Greenwich, RI. This Spill Plan
5
addresses the storage and containment of petroleum, oils, and lubricant products at
Camp Fogarty, and describes the “practices, procedures, structures, and equipment for
the prevention of and response to spills” that can potentially occur on the property
(RIARNG). The Spill Plan contains a number of reasonable spill scenarios, a section on
spill response procedures, and several maps of the various parking lots in the Camp that
contain aboveground storage tanks. These maps, which were created by an outside
contractor, contain estimated flow paths for pollutants that could potentially be spilled
within the parking lot. Unfortunately, these maps are not the product of a scientific
investigation, and as such, their accuracy is somewhat questionable. Accurate flow path
maps are especially critical for this location, as there are a number of small streams on
and around the property that drain to the Hunt River Watershed, and an improperly
contained petrochemical spill could prove disastrous for the local flora and fauna. This
project was begun to address this problem, and provide the RI Army National Guard
with science‐based flow path maps developed through hydrologic modeling in GIS.
II. Goal of Project
The ultimate goal of this project is to provide the Rhode Island Army National
Guard with a series of maps that depict the flow paths of liquid pollutants in two of their
parking lots, as well as an overall model of flow paths for Camp Fogarty, and a step‐by‐
step instructional manual that will guide users through the process of creating a flow
path map. This project is essentially a test to determine if it is possible to create
overland flow paths for chemical spills using hydrologic modeling tools within the
framework of a Geographic Information System. The parking lots adjacent to the Sun
Valley Armory and the Camp Fogarty Armory are the two areas that the flow path
analysis concentrates on. During the flow path analysis, LIDAR‐derived elevation data
are run through a hydrologic modeling extension of ArcGIS, flow direction and
accumulation are calculated, depressions and sinks are located, and flow paths are
created through the process of defining streams within the landscape. The maps that
result from this analysis will not only feature general flow paths for each parking area,
6
but also specific point source flow paths for each aboveground storage tank within the
parking lot, and the depressions within the parking areas that each flow path drains to.
III. Sources of Data
All of the data used in this analysis were supplied to the author in an ESRI
(Environmental Systems Research Institute, Redlands, CA) file geodatabase by Michael
Bradley and Tracey Daley of the Rhode Island Army National Guard. The primary vector
data that were used in the analysis include three feature classes: Vehicle Parking Areas,
Existing Structures, and Aboveground Storage Tanks. These feature classes are
represented in a WGS 1984 UTM Zone 19N coordinate system. The Aboveground
Storage Tank feature class is a point data set that was created in 2005 by georeferencing
the locations of aboveground storage tanks in Camp Fogarty using a Trimble GeoXT GPS
handheld unit. The Existing Structures feature class is a polygon data set that was
created in 2006 by georeferencing existing structures in Camp Fogarty from statewide
aerial imagery for 2003‐2004 provided by the RI Department of Transportation.
Similarly, the Vehicle Parking Areas feature class was created in 2010 by georeferencing
existing parking lots in Camp Fogarty from the 2008 RI Enhanced 911 statewide aerial
images provided by the firm Pictometry.
The primary raster data used in this analysis were a 1‐meter spatial resolution
Digital Elevation Model that was originally derived from LIDAR data. Digital Elevation
Models, or DEMs, are defined as “a raster set of elevations, usually spaced in a uniform
horizontal grid” (Bolstad 2008). EarthData International collected the original LIDAR data
in November 2006 at the request of the RI Army National Guard. Upon receipt of the
ASCII file format Bare Earth Grid, Tracey Daley converted them into text files, which
were then converted into ASCII 3D file format. Finally, the ASCII 3D file was converted
into a Bare Earth DEM. This 3 meter DEM was later resampled by Michael Bradley to a 1
meter DEM, the raster which is used in this analysis. This DEM, Fogarty_1m, is a
continuous floating point raster that is 1219 by 1739 in size (pixel rows and columns)
with a WGS 1984 UTM Zone 19N coordinate system.
7
IV. Methods
The following steps were used to conduct a flow path analysis for two parking
lots located within the RI Army National Guard property at Camp Fogarty. The analysis
conducted within this paper is performed entirely within ESRI’s ArcGIS 10 (Service Pack
3). ArcGIS is an extremely powerful and versatile geospatial data analysis and mapping
software package that is widely used in the field of environmental science. This analysis
largely relies on a free extension of ArcGIS, called ArcHydro, which is primarily used for
watershed mapping and delineation. I have outlined the steps of this analysis from start
to finish, beginning with downloading and installing ArcHydro, and ending with creating
custom point source flow paths for your map. Each step in the analysis begins with an
underlined heading, and a brief description of the analysis that you are about to
complete. Any text fields that you need to complete, buttons that you need to click,
menus that you need to select a choice from, or actions that you need to take will be
highlighted in bold. Additionally, important tasks or anything that you should be careful
with will be labeled with the prefix “NOTE:” and will be italicized.
Downloading Arc Hydro
Using the FTP client of your choice, FTP into the ESRI site and download the ArcHydro
installer.
• ArcHydro can be accessed using the following information: (Dartiguenave 2008)
o Server Address: ftp.esri.com
o Login: RiverHydraulics
o Password: river.1114
• Once inside, browse to ArcHydro > Setup10 > 2.0.1.133_2.0_Final
o Download ArcHydroTools.msi to your Desktop. The installer is a 21.8 mb
file, and takes only a few moments to download over a high‐speed
Internet connection.
8
Install ArcHydro
Next, using the ArcHydroTools.msi Installer, you will install the ArcHydro toolbar to your
computer.
• NOTE: Before you install ArcHydro, check to make sure that the following
software prerequisites are installed on your computer:
o Microsoft .Net Framework 3.5
o ArcGIS 10 with .Net libraries
o Spatial Analyst Extension
Note: While Spatial Analyst is not required to install ArcHydro, it
must be activated on your computer in order to run virtually all of
the tools used in this analysis.
• The ArcHydro installation is a fairly standard one, and makes use of the familiar
InstallShield software.
o Simply double click on the ArcHydroTools icon on your Desktop to begin
the installation process.
o Click Next , then Accept the Terms of the License Agreement, and
continue clicking on the Next button until the installation is complete.
o Click Finish to complete the installation.
Figure 1. ArcHydro Installation Wizard
9
• Finally, you will enable the ArcHydro toolbar within ArcMap.
o Open ArcMap, and right click in the toolbar area.
o Select ArcHydro Tools from the drop‐down menu. This will place
ArcHydro in your toolbar.
• Your installation is now complete, and you are ready to begin using ArcHydro.
Figure 2. ArcHydro Toolbar
Create A Geodatabase
To begin the process, create a new file geodatabase using ArcCatalog. This geodatabase
will act as a repository for the majority of the files that you create while performing a
flow path analysis.
• New > File Geodatabase
o In keeping with basic file‐naming protocol for ArcGIS, be sure not to use
any spaces or non‐alpha numeric characters other than dash or
underscore when naming your geodatabase.
o Name your geodatabase something meaningful; I suggest naming it after
your study area.
• Once your geodatabase is created, right‐click in the same folder where it is
located and create a New Folder.
o Name your folder the same thing as your geodatabase. You will use it
later to hold output from ArcHydro.
Select Your Study Area
Next, open a new map document, and add the DEM and the polygon feature class or
shapefile that contains your study area. If necessary, select out the specific location
where you would like to perform a flow path analysis.
10
• Selection > Select by Attributes
o Right click on the selected layer > Data > Export Data
o Browse to the location of your geodatabase, so that your selection will be
exported as a new feature class in your geodatabase
o Be sure to name your new feature class something meaningful. Ex:
Study_Area
Setting Up Your Geoprocessing Environment
When beginning any raster analyses, it is wise to first set up your geoprocessing
environment.
• Geoprocessing > Environment
• Workspace > Current Workspace: Your Geodatabase
• Output Coordinates > Output Coordinate System: Same as Input; WGS 1984
UTM Zone 19N
• Processing Extent > Extent: Same as clipped study area
• Raster Analysis > Cell Size: Same as DEM
• Raster Analysis > Mask: Clipped study area
• Raster Storage > Pyramid: Check Build Pyramids
Clip Your DEM
Before you can create a flow path map using ArcHydro, it is necessary to first clip your
DEM to the extent of the area you would like to perform your flow path analysis on.
• Add your DEM to the map document.
• ArcToolbox > Data Management Tools > Raster > Raster Processing > Clip
o Input Raster: The DEM you want to clip
o Output Extent: Your Study Area feature class
o Be sure to check the Use Input Features for Clipping Geometry check box
o Output Raster Dataset: Create a new folder for your clipped DEM, and
give it a name. The name of your clipped DEM is restricted to 13
11
characters or less. Do not save your new DEM to the geodatabase,
because the next step requires editing it outside of a geodatabase.
Instead, save it to the folder that you created earlier.
o Note: When specifying your output, you want it in ESRI Grid format, so
leave the file extension blank. This is very important to the next step! If
you create your clipped DEM in a format other than ESRI Grid, you will run
into problems that you cannot fix.
o Note: ArcHydro is not capable of processing grids larger than 20,000 x
20,000 cells.
• Press OK to run the command.
• Once the Clip procedure is finished, Save your map document, and exit ArcMap.
Enable Z‐Units in Your DEM
Unfortunately, the clipped DEM you have just created must be manipulated a little
before you can use ArcHydro. Regardless of if z‐units are enabled in your original DEM,
they are disabled by default in your clip, and must be manually enabled to take
advantage of the full functionality of ArcHydro.
• Browse to the location of your clipped DEM in Windows Explorer, and find the
prj.adf file associated with it
• Right click on the prj.adf file, and open it in a plain text editor of your choice.
Notepad, the Windows default will work fine for this.
Take note of the line of text that makes up this file, and find where it says "Zunit
NO".
• Replace "NO" with "1" (without the quotation marks), and save the file.
• Note: 1 represents the number of z units present in on spatial reference unit. If
the DEM's values were in meters, as they are in this example, you would put 1. If
the DEM's values were in centimeters, you would put 100 (Murison 2011).
• Be careful not to save the file in .txt format.
12
Data Management in ArcHydro
Before you begin work in ArcHydro, you must first establish how you want it to manage
data output. Unlike the tools available in ArcToolbox, ArcHydro does not give you the
option of selecting where you want your output to go for each individual command.
Instead, you must select one geodatabase to put all of your vector data in, and one
folder to put all of your raster data in.
• ArcHydro > ApUtilities > Set Target Location
o Raster Data: Browse to the folder you created previously, and set it as
the target location for your raster data.
o Vector Data: Browse to the geodatabase you created previously, and set
is as the target location for your vector data.
o Press OK to confirm your selections.
Figure 3. Data Management in ArcHydro.
Depression Evaluation
The Depression Evaluation Tool defines and characterizes depressions within a DEM, as
well as the areas that drain into depressions. The output of the Depression Evaluation
comes in two forms: the depressions themselves and the areas that drain into them.
Characteristics such as the volume and depth of each depression are generated by
default in the unit of measurement.
• ArcHydro > Terrain Preprocessing > DEM Manipulation > Depression Evaluation
13
• DEM: Select your clipped DEM from the drop‐down menu.
Depression: Name your depression output. Ex: ExampleDep.
Depression Drainage Area: Name your depression drainage area output. Ex:
ExampleDepDA.
Figure 4. Depression Evaluation Dialog
• Press OK to run the tool.
Depression Cleanup
The output that results from the depression evaluation tool has a very pixilated
appearance, and in general is aesthetically displeasing. To make this output suitable for
use in a map, it is necessary to smooth the polygons out.
• ArcToolbox > Cartography Tools > Generalization > Smooth Polygon
o Input Features: Select your Depression Evaluation depression output
from the drop‐down menu. Ex: ExampleDep
o Output Feature Class: Browse to the geodatabase you created
previously, and name your output. Ex: SmoothedDepression.
o Smoothing Algorithm: PAEK
o Smoothing Tolerance: 5 Meters
o Press OK to run the tool.
14
Figure 5. Original Output from Depression Evaluation
Figure 6. Depression Evaluation Output After Smoothing
Flow Direction
The flow direction tool analyzes the values of the 8 cells that surround a given cell and
determines the direction of the flow based on the steepest path (Czekanski & McKinney
2008).
• ArcHydro > Terrain Preprocessing > Flow Direction
15
o Hydro DEM: Select your clipped DEM from the drop‐down menu.
o Outer Wall Polygon: Select the clipped vector dataset containing your
study area from the drop‐down menu.
o Flow Direction Grid: Name your new flow direction raster output. Your
file name cannot exceed 13 characters, and as usual, should not contain
any spaces or non‐alphanumeric characters.
Figure 7. Flow Direction Dialog
• Press OK to run the tool.
• Your output should resemble the output below in Figure 8.
16
Figure 8. Flow Direction Output
Flow Accumulation
The flow accumulation tool computes a raster grid that contains the accumulated
number of cells upstream of a given cell. Flow accumulation is the basis for the flow
path analysis.
• ArcHydro > Terrain Preprocessing > Flow Accumulation
o Flow Direction Grid: Select the flow direction raster that you just created
from the drop‐down menu.
o Flow Accumulation Grid: Name your new accumulation grid. As this is a
raster file, your file name cannot exceed 13 characters.
o Press OK to run the tool.
Figure 9. Flow Accumulation Dialog
• Your output will resemble the output below in Figure 10.
17
Figure 10. Flow Accumulation Output
Stream Definition & Segmentation
• ArcHydro > Terrain Preprocessing > Stream Definition
o A warning will pop up telling you to enter a numeric value greater than zero
for the Area. This number represents the Stream Threshold value. Press OK
and ignore it for the time being, leaving it at zero. If the output appears to be
too dense (i.e., too many streams), this is the number to change. A smaller
number will result in a denser stream network, while a larger number will
result in a less dense stream network (Czekanski & McKinney 2006).
o Flow Accumulation Grid: Select your flow accumulation raster from the drop‐
down menu.
o Stream Grid: Name your new stream grid. Because this is a raster file, your
name cannot exceed 13 characters.
18
Figure 11. Stream Definition Dialog
o Press OK to run the tool.
• ArcHydro > Terrain Preprocessing > Stream Segmentation
o Flow Direction Grid: Select your flow direction raster from the drop‐down.
o Stream Grid: Select the stream definition grid that you just created.
o You may leave Sink Watershed Grid and Sink Link Grid set to Null, as these
datasets have not been created and are not necessary to this step.
o Stream Link Grid: Name your new stream link grid. This is a raster, so your file
name cannot exceed 13 characters.
19
Figure 12. Stream Segmentation Dialog
Press OK to run the tool.
Drainage Line Processing
This is the process that actually creates the flow lines that will go into your flow path
map.
• ArcHydro > Terrain Preprocessing > Drainage Line Processing
o Stream Link Grid: Select the stream link grid created by the Stream
Segmentation tool from the drop‐down menu.
o Flow Direction Grid: Select your flow direction raster from the drop‐
down menu.
o Drainage Line: Name your new flow lines.
o Press OK to run the tool. This process may take a few minutes to
complete.
o Your output will resemble the output below in Figure 13.
o Once output for the Drainage Line Processing tool is generated, you have
all of the data you need to make a general flow path map for your study
area. If you would like to generate flow paths for specific points within
your study area, continue on to the next step.
20
Figure 13. Drainage Line Processing Output. This output is your finished flow path result.
Interactive Flow Path Tracing
One of the more interesting and useful tools available in ArcHydro is the Interactive
Flow Path Tracing tool. This tool allows the end user to create a custom flow path for
any point source within the boundaries of the DEM.
• ArcHydro > Interactive Flow Path Tracing
Figure 14. Interactive Flow Path Tracing on the ArcHydro Toolbar
o Drainage Area: Select the Drainage Area output created by the
Depression Evaluation Tool.
Note: Make sure to select the Drainage Area output, and not the
Depression output.
o Flow Direction Grid: Select your Flow Direction raster from the drop‐
down menu.
o Longest Flow Path: Name your interactive flow path output.
21
Figure 15. Interactive Flow Path Tracing Dialog
• Press OK to begin using the tool.
• To use this tool, simply click anywhere within the DEM, and ArcHydro will draw a
custom flow path in vector line format for the point where you’ve clicked.
If the flow path drains to an area outside the boundaries of the DEM, the tool
will return an error: No drainage area found for the clicked point! Click OK to
dismiss it.
• Otherwise, if your flow path drains to an area within your DEM and is successful,
you will automatically zoom to your new flow path and a dialog box will pop up
and ask you if you want to create a flow path as shown. Click Yes to approve
your new flow path.
• Flow paths generated by this tool will always automatically drain to the nearest
depression, regardless of the size or volume of the depression. If you believe
that the volume of liquid flowing along your custom path exceeds the capacity of
the depression it first drains to, simply create another flow path just beyond that
depression to determine where the liquid will flow next.
22
Figure 16. Interactive Flow Path Tracing Output. The squiggly red lines represent two custom flow paths
generated by the tool. Note that they automatically stop at the depression that they drain to.
V. Results
The results of the flow path analysis for the Sun Valley Armory parking lot were
as expected. The analysis produced extremely clear and well‐defined flow paths that
lead to depressions that can be observed on site with the naked eye. As indicated in
Figure 17, there is a large depression nearly the length of the parking lot that runs from
North to South. This depression is located approximately 35 feet to the West of the
motor pool gate, and the majority of the parking lot drains to this location. This
depression is capable of holding approximately 34,196 US Gallons of fluid.
23
Figure 17. Results of the Sun Valley Armory flow path analysis
To the East of this large depression, there are a number of smaller depressions, each
capable of holding between 3 and 2,375 US Gallons of fluid. Most of the eastern portion
of the parking lot will drain to these depressions first, only draining to the larger
depression if the volume of fluid spilled exceeds their capacity. The Interactive Flow
Path Tracer shows that each of the two aboveground storage tanks will drain to a
smaller depression, as indicated in Figure 18. The 500 Gallon storage tank will drain to a
3‐gallon depression before overflowing, and the 4000 Gallon storage tank will drain to a
21‐gallon depression before overflowing. Once these smaller depressions overflow, the
liquid from the tanks will travel west to the largest depression, where they will be
contained completely.
24
Figure 18. Interactive Flow Paths for the Sun Valley aboveground storage tanks and the surrounding area.
Unfortunately, the results of the flow path analysis for the Camp Fogarty Armory
parking lot were neither clear, nor particularly well defined. The first run‐through of the
analysis produced results that were completely unsuitable for any kind of use.
25
Figure #. Initial results of Camp Fogarty Armory flow path analysis. Stream threshold value = 0
These first flow paths covered nearly the entire study area, and had no discernable
pattern. By adjusting the Stream Threshold value in the Stream Definition process from
0.0000 to, it was possible to decrease the density of the stream network and produce a
slightly more defined set of flow paths. However, these new flow paths are too short
and choppy to be of much use. The flow paths do seem to indicate that they drain to
depressions in the landscape. There are a number of small depressions throughout the
parking lot, the largest of which is capable of holding approximately 519 US Gallons of
fluid. This depression is located in the southwest corner of the parking lot, where it
appears that the majority of spilled fluid will go.
26
Figure 19. Secondary Results of Camp Fogarty Armory flow path analysis.
Stream threshold value = 0.000010
The results produced by the Interactive Flow Path Tool are far clearer than those
produced by the general flow path analysis. The Interactive Flow Path Tool shows that
fluid leaked from the 3000 Gallon aboveground storage tank will drain to a series of
small depressions, finally stopping at a 108 gallon depression located approximately 150
feet to the southeast of the tank. Once this depression overflows, the fluid from it will
pool around it and remain in the southwest corner of the parking lot.
VI. Discussion & Conclusion
The combination of the success of the flow path analysis for the Sun Valley
Armory parking lot and the somewhat unexpected failure of the flow path analysis for
the Camp Fogarty Armory leads me to believe that while the model of creating flow
paths in ArcHydro is valid, the model is also strongly susceptible to problematic data. I
27
believe that the unfavorable flow paths created by the Camp Fogarty Armory analysis
are likely the result of artifacts in the original data. I do not believe that the results
generated by this flow path analysis occurred as a result of a problem with the actual
algorithms used by ArcHydro. When looking at the clipped DEM for the parking lot,
clearly visible diagonal striations are present throughout. These striations did not
appear to be present in the actual parking lot on the date of my last field visit, and I
strongly recommend that the Army National Guard conduct a survey of the site to
determine if they exist in the real world.
Figure 20. Potential artifacts in the DEM
It is possible that this problem may potentially be remedied through converting the
floating point raster to integer format. This conversion process has been known to
reduce the number of sinks and artifacts in a DEM. However, this process relies on
knowing the vertical accuracy of the data, something that has not yet been established
for this dataset and is currently unknown.
In light of the unclear and somewhat disappointing nature of these results, I
have created a workaround to generate flow paths for the Camp Fogarty Armory
parking lot. Looking at the output of the flow accumulation tool, it is clear that some
sort of fluid stream exists in the parking lot. By using Raster Calculator and Reclassify, it
is possible to isolate these accumulation streams so that the raster is comprised of
streams with a value of 1 and all other cells with a value of NoData. Next, using the tool
Raster to Polyline, it is possible to convert the raster accumulation streams to line
28
format, effectively creating flow lines for the parking lot. These streams do require some
editing, but the end result is a set of easily interpretable flow paths.
To ensure that these results have some validity, I carried out this procedure on
the flow accumulation output for the Sun Valley Armory parking lot and compared them
to the real flow paths generated by the Drainage Line Processing Tool; not surprisingly,
the results are nearly identical (Figure 21).
Figure 21. Comparison of results for Sun Valley Armory
Of course, it will be entirely up to the Army National Guard to decide which map to use
for the Camp Fogarty Armory parking lot.
29
Figure 22. Final Camp Fogarty Armory Map
Looking at the big picture, it is clear that ArcHydro is capable of producing
tangible flow paths for a given area, and that it has a valid place in the world of
predictive oil spill modeling. While ArcHydro cannot take into account factors such as
dispersal or dissolution, it can provide managers and scientists with basic information
about the trajectory an overland chemical spill can take. As with most things, the results
of a flow path analysis in ArcHydro are only as good as the data used as input. Problems
within the data will almost certainly cause problems with the model. ArcHydro is a very
powerful extension, with a tremendous amount of capabilities, and it should be noted
that this analysis has barely scratched the surface of what ArcHydro can actually do. The
RI Army National Guard will certainly be able to take advantage of these capabilities in
the future, and I strongly believe that the process used to create these flow paths can
be replicated for other National Guard properties. Should this process be replicated in
30
the future, I recommend that the Guard strive to obtain the highest spatial resolution
data possible, and that all raster data should be screened in advance for potential
artifacts or anything else that could negatively impact the model. Additionally, I would
recommend that the Guard perform empirical tests in their parking lots to confirm the
validity of future results. This would entail systematically releasing water at various
locations and observing flow paths to see if they follow the paths predicted in this
analysis. Empirical assessment of depression volumes might also be done, however this
process can be complicated by rapid soil or gravel infiltration of fluids, which can result
in higher capacities that what was predicted by elevation‐derived volumetric
estimations.
31
Acknowledgements
To begin, I’d like to thank Dr. Peter August for suggesting this project to me, and
for being an invaluable source of advice and encouragement throughout the major
paper process and my time as a MESM student. Additionally, I would like to thank
Michael Bradley and Tracey Daley of the RI Army National Guard for allowing me the
opportunity to work on this project, granting me access to this data, and providing
technical assistance. I would also like to thank my lab mates in the EDC Lab, Heather
Grybas and Tiffany‐Lane Davis, for their encouragement and emotional support. I’d also
like to thank the NRS Department for taking me on as the Department Webmaster, and
for allowing me the many wonderful opportunities I’ve had during my time in the MESM
program. Finally, I’d like to thank my husband Keith for supporting me throughout my
post‐graduate career; I couldn’t have done this without his constant encouragement
and help.
32
References
Adam, G. and H.J. Duncan. (1999) Effect of diesel fuel on growth of selected plant
species. Environmental Geochemistry and Health. 21: 353‐357.
Asia‐Pacific Applied Science Associates (APASA) (2003) A review of recent innovations
and current research in oil and chemical spill technology. Available online at:
http://www.amsa.gov.au/marine_environment_protection/national_plan/contin
gency_plans_and_management/research_development_and_technology/spill_t
echnology.pdf
Bolstad, P. (2008) GIS Fundamentals: A First Text on Geographic Information Systems.
Elder Press: White Lake, Minnesota.
Czekanski, A.J. and D.C. McKinney. (2006) Introduction to Arc‐Hydro: ACEH basin pilot
study. Center for Research in Water Resources: Austin, Texas. Available online
at: http://www.ce.utexas.edu/centers/crwr/reports/online.html.
Dartiguenave, C. (2010) ArcHydro in ArcGIS 10. [Message 6] Posted to:
http://forums.arcgis.com/threads/8274‐ArcHydro‐in‐ArcGIS‐10.
Ivanov, A.Y. and V.V. Zatyagalova. (2008) A GIS approach to mapping oil spills in a
marine environment. International Journal of Remote Sensing. 29, 21: 6297‐
6313.
Li, Y., A.J. Brimicombe, and M.P. Ralphs. (2000) Spatial data quality and sensitivity
analysis in GIS and environmental modelling: the case of coastal oil spills.
Computers, Environment and Urban Systems. 24: 95‐108.
Lloyd, A.C. and T.A. Cackette. (2001) Diesel Engines: Environmental Impact and Control.
Journal of the Air and Waste Management Association. 51: 809‐847.
Lytle, D.A. and B.L. Peckarsky. (2001) Spatial and temporal impacts of a diesel fuel spill
on stream invertebrates. Freshwater Biology. 46: 693‐704.
Mahaney, P.A. (1994) Effects of freshwater petroleum contamination on amphibian
hatching and metamorphosis. Environmental Toxicology and Chemistry. 13, 2:
259‐265.
33
Murison, L. (2011) Error in Level DEM: Spatial reference does not have z unit and in Sink
evaluation. [Message 2] Posted to: http://forums.arcgis.com/threads/34621‐
Error‐in‐quot‐Level‐DEM‐quot‐Spatial‐reference‐does‐not‐have‐z‐unit‐and‐in‐
quot‐Sink‐evaluation‐quot.
Rhode Island Army National Guard. (2010) East Greenwich Camp Fogarty Training Site
Spill Prevention and Contingency Plan.
Sorenson, M. (1995) Arc/INFO Marine Spill GIS. Spill Science & Technology Bulletin. 2, 1:
81‐85.
Vandermuelen, J.H. and C.W. Ross. (1995) Oil Spill Response in Freshwater: Assessment
of the Impact of Cleanup as a Management Tool. Journal of Environmental
Management. 44: 297‐308.