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
33
Embed
Predictive Modeling of Overland Spill Using ArcGIS M. · Predictive Modeling of Overland Petrochemical Spill Trajectories Using ArcGIS by Kristen M. Mathieu A MAJOR PAPER SUBMITTED
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
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.