MRiea REPORT
Specifications for Street Surface Loading Model
For U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
EPA Contract No. 68-DO-0123
Work Assignment No. II-77
MRI Project No. 9712-M(77)
April 20, 1993
MIDWEST RESEARCH INSTITUTE 425 Volker Boulevard, Kansas City, MO 64110-2299 • (816) 753-7600
-------
REPORT
Specifications for Street Surface Loading Model
For U.S. Environmental Protection Agency
Office of Air Quality Planning and Standards
Research Triangle Park, NC 27711
Attn: Scott Voorhees
Air Quality Management Division (MD-15)
EPA Contract No. 68-DO-0123
Work Assignment No. II-77
MRI Project No. 9712-M(77)
April 20, 1993
MIDWEST RESEARCH INSTITUTE 425 Volker Boulevard, Kansas City, MO 64110-2299 • (816) 753-7600
-------
PREFACE
This Software Specifications Document was prepared for Mr. Scott Voorhees of
the SOj/Particulate Matter Programs Branch, Air Quality Management Division, Office
of Air Quality Planning and Standards, U.S. Environmental Protection Agency,
Research Triangle Park, North Carolina, under EPA Contract No. 68-DO-0123, Work
Assignment No. 77. The Document describes the scope, specifications, and initial
design of the computer model for street surface loadings. MRI's Project Leader for
the assignment is Mrs. Mary Ann Grelinger.
MIDWEST RESEARCH INSTITUTE
ird V. Crume
Program Manager
Environmental Engineering Department
Approved:
Charles F. Holt, Ph.D., Director
0 Engineering and Environmental
Technology Department
April 20, 1993
MRI-MVR9712-77.12
III
-------
CONTENTS
Preface iii
Figues vii
Tables viii
1. Introduction 1
1.1 Purpose of software 1
1.2 Background 2
1.3 Organization of document 2
2. System Characteristics and Requirements 3
2.1 Conceptual model of silt loading dynamics 4
2.2 Hardware and software requirements 5
2.3 Software documentation 5
3. Model Algorithms 7
3.1 Basic model 7
3.2 Model parameters and relationships 10
3.3 Model pseudo-code 13
3.4 Program sequence 17
4. Data Structures 21
4.1 Program data structures 21
4.2 Results data base 24
4.3 Program output 24
4.4 Auxiliary data bases 24
5. Model Development 27
6. References 29
Appendices
A Input screen displays A-1
B Analysis of pseudo-random numbers generated using
Foxpro® routines B-1
C Silt loading data base (from WA No. 44, EPA Contract
No. 68-DO-0123) C-1
MRI-M\R9712-77.12
-------
LIST OF FIGURES
Number Page
3-1. Model pseudo-code 14
4-1. Output file structure 25
4-2. Report format (one page for each road class) 26
MRI-MVR9712-77.12 VII
-------
LIST OF TABLES
Number Page
3-1. Parameters to be modeled 8
4-1. PARM data formats , 22
5-1. Schedule of deliverables 27
MRI-M\H9712-77.12 VIM
-------
SPECIFICATIONS FOR STREET SURFACE LOADING MODEL
SECTION 1
INTRODUCTION
This document provides an understanding and background of the Environmental
Protection Agency's (EPA) specification for a mathematical model to provide more
reliable estimates of paved road surface silt loading. Measured as mass of material
per unit area, silt loading is the major parameter associated with estimation of PM10
emissions from paved roads using EPA's emission factor equation presented in AP-42,
Section 11.2. PM10 is defined as airborne paniculate matter with aerodynamic
diameters equal to or less than 10 u,m. Silt is the fraction of particles with physical
diameters less than 75 u.m, as determined using a dry sieving methodology.
This report responds to the Work Assignment requirement that "all software
development shall conform with EPA's Office of Information Resources Management
(OIRM) policy requirements as specified in the System Design and Development
Guidance, Volumes A, B, and C. (OIRM, June 1989). The scope, specifications, and
initial design of the computer model are also discussed in this report.
1.1 PURPOSE OF SOFTWARE
The purpose of this computer model is to simulate the dynamic changes in
paved road surface loading over time and space, taking into account variable deposi-
tion and removal processes. The goal of this simulation will be to reduce the need for
costly sampling and analysis procedures to quantify particulate emissions from paved
roads. A computer model will allow more efficient characterization of paved road silt
loading for development of PM10 emission inventories by state and local agencies.
The model will provide the end user a tool with which to quickly and efficiently obtain
representative silt loading values for the variety of paved roads in a metropolitan area.
The computer model will estimate silt loadings for specific road classes, geographic
regions and seasons, and will be made available to state and local agencies.
MRI-M\R9712-77.12 1
-------
1.2 BACKGROUND
Midwest Research Institute has provided considerable support to the EPA in the
area of measurement and control of particulate matter emissions. A partial listing of
documents produced by MRI researchers may be seen in Section 6. Many of these
reports describe the collection of dust samples (and analysis for silt content) taken
from paved road surfaces, and also discuss the relationships between paved road silt
loading and other parameters, such as average daily traffic (ADT) and tire track-on of
dirt from unpaved areas. These and other documents have produced input data and
algorithms to be incorporated into the computer model.
1.3 ORGANIZATION OF DOCUMENT
Section 2 covers the specifications and design of the paved road surface silt
loading model. Included is a discussion of the operation, input requirements, and
limitations of the model. The hardware platform and software tools are specified in
Section 2.
Section 3 presents the conceptual model to estimate street surface silt loadings,
and will describe the algorithms and data structures associated with the design of the
model.
Section 4 discusses the quality assurance aspects of the software development.
Section 5 concerns documentation of the software.
Appendix A displays preliminary model input screens.
Appendix B presents an analysis of pseudo-random numbers generated by the
FoxPro® model program.
Appendix C shows the silt loading data base with measurement data collected
by MRI to support emission factors for paved roads presented in AP-42,
Sections 11.2.5 and 11.2.6.
MRI-M\R9712-77.12
-------
SECTION 2
SYSTEM CHARACTERISTICS AND REQUIREMENTS
EPA, in the Scope of Work section of the Work Assignment, has defined the
critical parameters to be considered for inclusion in the surface silt loading model. At
a minimum, the computer model will incorporate dust removal factors such as street
sweeping, precipitation, and vehicular traffic. It will also incorporate dust deposition
factors such as pavement wear, tire wear, material spilled from truck loads, trackout
and washout from surrounding unpaved areas, and sanding/salting for snow/ice con-
trol. Other factors that will be considered for inclusion into the model are geographic
location, land use, season, and type of road surface. In addtion, meteorology (temper-
ature and rainfall) and traffic density will be included. The computer program will
model events that are known to both augment and decrease silt loadings on a paved
road.
A stochastic (pseudo-random) model is planned to enable a more realistic
simulation of the varying nature of silt loading on a road from day to day as compared
to a deterministic model. The advantage of a stochastic simulation is that parameters
can be varied to determine the sensitivity of surface silt loading to each model para-
meter. A simulation run will consist of modeling at least 1,000 seasonal days to
create a matrix of silt loadings for each road class together with the model para-
meters. If the parameters affecting silt loading are correctly included in the simulation,
and are adequately described by the statistical distribution assigned to each para-
meter, the results should be representative of actual silt loading variation for the
particular season and road class being modeled.
MRI-M\R9712-77.12
-------
2.1 CONCEPTUAL MODEL OF SILT LOADING DYNAMICS
The background silt loading on a paved roadway represents a balance between
the rate of deposition and the rate of particle entrainment (emissions). The back-
ground (equilibrium) silt loading is inversely dependent on average daily traffic (ADT).
For example, freeways have a much lower equilibrium silt loading than collector roads.
When a high deposition event occurs (e.g., a spill), the equilibrium is
temporarily upset and the restoration process begins. Assuming that the spill is
represented by a step increase in silt loading over a unit length of roadway, the rate of
return to equilibrium can be calculated because the emission rate is a known function
of silt loading, as experimentally determined from past MRI field tests.
The emission rate starts from its highest value immediately after the spill and
decreases as the silt loading decreases. The background deposition rate (equal to the
emission rate at the equilibrium loading) continues throughout the restoration period.
The rate of return to equilibrium depends on the ADT.
If a first order dependence of emissions on silt loading is assumed, the elevated
loading will approach the equilibrium value asymptotically following an exponential
decay function. Therefore, a certain number of half-lives might be considered a full
return to equilibrium.
When a sudden silt loading removal event occurs (e.g., a street flushing), there
is a step decrease in silt loading and the restoration process begins. The background
deposition rate exceeds the entrainment (emission) rate during the restoration
process. Again the return to equilibrium can be represented mathematically by
making use of the known dependence of emissions on silt loading.
Mud and dirt track-on is the most important cause of elevated silt loadings in
urban areas across the country. MRI has obtained some field data showing the
distribution of silt loading around the peak values at the point of access to the source
of track-on (e.g., a construction site). Under relatively dry conditions, dirt track-on
rates should be fairly consistent and predictable based on the number and type of
vehicles exiting the site onto the paved roadway. The general functional form of the
MRI-MVR9712-77.12
-------
spatial distribution of silt loading should provide valuable information both on the
track-on rate and the new "equilibrium" condition that is created.
2.2 HARDWARE AND SOFTWARE REQUIREMENTS
The PAved Road surface silt loading Model (FARM) will be developed on and
for an MS-DOS personal computer (PC) with an Intel 80386 or compatible micropro-
cessor. EPA is using IBM compatible PCs running MS-DOS applications for many of
its models. The advantages to this hardware platform are many, especially a common
architecture and operating system across many vendors and with a wide installation
base, cost-effectiveness, and the availability of a great variety of support software
(both for program development and for secondary data processing). The minimum
architecture will be utilized, with no expanded memory.
FoxPro® 2.x will be employed as the development software and will be
compatible with other EPA models and data base management systems, including
TANKS, SAMS, AMS-PC, etc. This data base management software enables the
easy use and manipulation of support data bases for program development. A
runtime program will be produced that will not require additional software acquisition
by the user.
The PARM model will be distributed to users of EPA's Technology Transfer
Network (TTN) bulletin board system (BBS) operated by the Office of Air Quality
Planning and Standards (OAQPS). PARM will be placed on the Clearinghouse for
Inventories and Emission Factors (CHIEF) BBS, which provides access to tools for
estimating emissions of air pollutants and performing air emission inventories.
2.3 SOFTWARE DOCUMENTATION
Documentation of the" model will be provided to users as a WordPerfect 5.1 file
suitable for distribution to users via the CHIEF BBS. Copies will be supplied to EPA
as required in the Work Assignment. The documentation will discuss the rationale for
decisions made in developing the model and also describe the following:
• Data entry and update processes
MRI-MVR9712-77.12
-------
Internal programs and algorithms
Program code and data bases
Ad-hoc capabilities and program run options
Reporting options
Help facilities and tools including on-line Help {F1}
Explanation of error messages
Auxiliary data base structures
User support and problem reporting
MRI-MVR9712-77.12
-------
SECTION 3
MODEL ALGORITHMS
3.1 BASIC MODEL
The computer program to model paved road surface silt loading over time will
be based on information contained in previous MRI documents and in other references
dealing with paved road surface loadings. Deposition and removal of silt (measured
as mass of material per unit area) on paved roads are known to be correlated with
vehicle types, vehicle traffic counts, road curbs, rainfall events, road classes,
geographic regions, seasons, etc. The following discussion introduces a preliminary
estimation of these relationships, which if previously unqualified, will be estimated
from underlying physical phenomena.
Only major silt deposition and removal events known to affect street surface
loadings will be quantified in the model. The following equation summarizes the
algorithms to be used in the model:
sL, = sUi + Du - Rg
where Rl} < (sLM + D^)
or sl_j > 0
where: sLi = Silt loading on paved road surface at end of day, i
sLM = Silt loading at end of previous day, i-1
Dj j = Deposition amount from activity j on day i
RJJ = Removal amount from activity j on day i
MRI-MXR9712-77.12
-------
Table 3-1 presents the variables to be modeled in the simulation, together with
their name abbreviations, whether they are expected to occur "E'Very day or are an
"Occasional event, their estimated occurrence pattern or dependency on other
variables, and the estimated statistical distribution of the variable.
TABLE 3-1. PARAMETERS TO BE MODELED
Variable
Silt loading
Background silt
loading
Salting/Sanding
Spills
Track-on
Wash-on
Traffic (ADT)
Precipitation
Street cleaning
Street litter
Pavement wear
Tire wear
Wind speed
Atmospheric
disposition
Abbr
SL
SLB
DG
DL
DC
DW
RF
RP
RC
DT
DP
DR
WS
DA
E/O
E
O
O
0
O
0
E
O
O
E
E
E
E
E
Occurrence pattern or
dependency
As simulated
From data base of silt
loadings
Season and rainfall
Traffic and spill
characteristics
Rainfall amount, road
class and land use
Rainfall amount, road
class and land use
Road class and land
use
Whether it rains and
how much
Season, sanding/
salting, carry-out,
wash-on, and land use
Season and land use
Traffic and road
surface type
Traffic and road
surface type
Region and season
Region, season and
wind speed
Distribution of
deposition/
removal amount
Unknown
Empirically determined
Empirically determined
Normal
Lognormal
Lognormal
Normal
First order Markov
process and Gamma
Poisson
Normal
Normal
Normal
Type I (G umbel) for dry
days only
Lognormal
E/O—E_very day or Occasional event
MRI-MW9712-77.12
8
-------
Several parameters in Table 4-1 are believed to contribute little to the
estimation of silt loading on a paved street. These include atmospheric deposition,
wind speed and street litter. Winds are known to produce high background levels of
fugitive dust, causing more coarse dust deposition on the road but also sweeping the
road clean of fine dust. Street litter consists mainly of larger materials such as paper,
leaves, etc., and is estimated to have little effect on the silt fraction.
The program will incorporate a simulation of daily rainfall and temperature that
affect several silt-loading parameters and will be suited for various geographical
regions and seasons.
The mathematical relationships that will be utilized include:
• Relationship of ADT to the removal of silt from a road.
• Frequency and distribution of silt quantities spilled by trucks (by road
classes, ADT, etc.).
• Frequency and distribution of silt quantities deposited by track-on (by
ADT, road class, etc.).
• Frequency and distribution of silt quantities deposited by wash-on (by
road class).
»
• Frequency and distribution of sanding/salting quantities deposited for
snow/ice control (by road class, temperature, etc.).
If correlations cannot be quantitatively defined based on empirical data,
estimates of the range of silt loadings deposited or removed by various activities will
be prepared based on MRI experienced judgment and known physical phenomena.
The model will assume a basic road segment, consisting of 1-mi by 1-lane, for
each road classification. All deposition and removal events will be characterized by
the quantity of silt (< 75 urn diameter) added or removed from the paved road surface.
The mean silt loading on each 1-mile road segment to be simulated by the model will
be assumed to apply to x miles of specified roadway within the city. The activity level
MRI-M\R9712-77.12
-------
used to calculate emissions is to be determined from the roadway mileage totals by
road classification (and associated ADT). Basic model parameters and associated
algorithms are described below.
3.2 MODEL PARAMETERS AND RELATIONSHIPS
Most model parameters will be stored in a data base, SIM_PARM.DBF, that
can be changed by the user. These parameters, assumptions, and statistical
distributions are discussed below.
Daily Rainfall
Daily rainfall will be simulated using a first order Markov process. The
probability of rainfall is dependent on a variety of meteorological parameters. A first
order Markov process is useful in describing the probability of rainfall as conditionally
dependent on whether it rained yesterday. This simplification allows rainfall (or dry
spells) events to be simulated in extended daily sequences that commonly happen in
nature. A 2 x 2 table can be constructed for rainfall probabilities, given that it did (not)
rain yesterday.
PI.I
P2.1
Pi*
P2.2
In addition, rainfall amounts can be generated using a gamma distribution.
Statistical parameters of rainfall occurrence and amounts will be determined from an
analysis of approximately ten years of Local Climatological Data associated with
several U.S. cities. Processed meteorological data will be stored in the computer data
base, METEOR.DBF, for access by the program.
The possibility of precipitation will be considered for each day. If it rains, a
precipitation amount will be generated. If precipitation is accompanied by sub-freezing
temperatures, then it is assumed to be snowing, and a sand/salt amount will be
calculated.
MRI-M\R9712-77.12 1 0
-------
Wash-on and track-on events are highly dependent on rainfall, and a history of
rainfall events will be kept for at least 3 days because surrounding bare, unvegetated
areas may remain muddy for several days. If a street segment is identified as
uncurbed, track-on will be simulated.
Track-on of Mud/Dirt onto Paved Roads from Unpaved Areas
Mud and dirt can be carried onto a paved surface by vehicles that enter the
roadway from unpaved areas. In this instance, the user may modify the default
frequency of occurrence of track-on. The amount of dirt tracked onto the road is
dependent on (a) the number of wheels passing from the unpaved area to the paved
road and (b) the surface characteristics of the site from which the dirt was tracked. In
this model, a typical vehicle producing track-on is assumed to be a 6-wheel truck
entering the roadway from a construction site. Only the silt fraction of the tracked-on
dirt will be considered. In the default operation of this model, track-on is assumed to
occur only on collector and local roads with no curbs. The user may modify default
values in the file SIM_PARM.DBF.
Wash-on of Mud/Dirt onto Paved Road during Heavy Rain Events
The frequency of wash-on of mud and dirt onto a typical local paved road
segment must be characterized by the local agency. A default wash-on value will be
available and will apply only to local and collector roads. The statistical distribution
most representative of this amount is estimated to be a normal distribution with a
default relative standard deviation (RSD) of 20%. An agency may choose to modify
all default numbers in the SIM_PARM data base.
Sanding and Salting
Sanding and salting of roads will be modeled in the program as occurring after
it rains and when the mean temperature is less than 32°F. Sand/Salt will be assumed
to be applied at a default rate of 500 Ib/lane mile for all road classes during a snow/ice
storm, and with a sand silt content of 3%. A normal distribution and RSD of 20% will
be assumed. These values may be modified by the user in the SIM_PARM data
base.
MRI-M\R9712-77.12
11
-------
Spill Events
Spill events will be modelled using a default mean silt amount deposited by a
certain fraction of heavy-duty trucks. All roadway classes are assumed to be evenly
affected.
Street Sweeping
Wet street sweeping will be assumed to clean a roadway of the silt loading
according to an efficiency stored in the SIM_PARM.dbf file. A default 96% efficiency
may be changed by the user. The frequency of wet street sweeping is stored in the
RDCLASS data base by road class and must be entered by the user.
Average Daily Traffic (APT)
ADT will be characterized by the type of vehicle traffic for each road class. In
this simulation, default 1990 vehicle mix fractions from the EPA MOBILE 4.0 model
will be used.
Light-duty gasoline vehicles (LDGV) 0.710
Light-duty gasoline trucks (LDGT) 0.213
Heavy-duty gasoline trucks (HDGT) 0.015
Light-duty diesel vehicles (LDDV) 0.013
Light-duty diesel trucks (LDDT) 0.004
Heavy-duty diesel vehicles (HDDV) 0.034
Motorcycles (MC) 0.010
ADT values are measured for each road class by local/state departments of
transportation and should be used in the simulation to replace default values in the
RDCLASS data base. The removal of silt leading from a road segment (1 lane mile)
will be expressed as a fraction of ADT.
MRI-M\R9712-77.12 12
-------
Tire. Brake, and Pavement Wear
Other model parameters affected by ADT are tire wear, brake wear, and
pavement wear. The EPA has developed suitable PM10 emission factors for these
three components of the model.
Tire Wear (light-duty vehicles)
E10 = 2 mg/VMT
Brake Wear (light-duty vehicles)
E10= 13 mg/VMT
Pavement Wear
E = ?
3.3 MODEL PSEUDO-CODE
The computer model will be based on the linear program structure shown in
Figure 3-1.
Initialization of the program will include specifications of nearest city, road
segments (functional road classes), season, and statistical distribution parameters.
In the activity sequence portion, the modeling program will generate pseudo-
random variates in a pseudo-random sequence for each day. "Cleaning, "Deposition
and "FTainfall events are each considered to generate six (3!) possible activity
sequences: CDR, CRD, RCD, RDC, DRC, and OCR, as previous described. All
cleaning events will be treated together in the sequence and all deposition events also
will be included together. These program sequences are described below in more
detail.
MRI-M\R9712-77.12
13
-------
Figure 3-1. Model pseudo-code.
begin
modify simulation parameters (As needed)
while DayCount < N
generate activseq (Daily order of rainfall, cleaning and deposition events)
generate rainfall event (Y/N and quantity), temperatures
for each road class
generate ADT, spill (Y/N and quantity), track-on (Y/N and quantity),
wash-on (Y/N and quantity), street cleaning (Y/N)
CASE1: Order CDR
Subtract from SL due to ADT cleaning
If street cleaning, then SL = SL * (1 - CE) (function of ADT,
road class, and land use)
Add to SL due to pavement wear (function of ADT)
Add to SL due to tire wear (function of ADT)
If spill, then add to SL (function of ADT, road class, and land
use)
Add to SL due to dry track-on (function of ADT, road class, and
land use)
Add to SL due to mud track-on (function of ADT, road
class, and land use; only if previous 3-day rainfall;
current day's rainfall not considered)
If rain event, then SL = SL/2
Save SLs and other parameters of day
CASE 2: Order OCR
Add to SL due to pavement wear
Add to SL due to tire wear
If spill, then add to SL
Add to SL due to dry track-on
Add to SL due to mud track-on (dependent only on previous
days' rainfall, not to current day's rainfall)
Subtract from SL due to ADT cleaning
If street cleaning, then SL = SL * (1 - CE)
If rain event, then SL = SL/2
Save SLs and other parameters of day
MRI-M\R9712-77.12 1 4
-------
Figure 3-1 (Continued)
CASE 3: Order CRD
Subtract from SL due to ADT cleaning
If street cleaning, then SL = SL * (1 - CE)
If rain event, then SL = 0
Add to SL due to pavement wear
Add to SL due to tire wear
If spill, then add to SL
Add to SL due to dry track-on
Add to SL due to mud track-on
Add to SL due to mud wash-on (only for collector and local
roads)
Add to SL due to sand/salt (dependent on rainfall and
temperature)
Save SLs and other parameters of day
CASE 4: Order DRC
Add to SL due to pavement wear
Add to SL due to tire wear
If spill, then add to SL
Add to SL due to dry track-on
Add to SL due to mud track-on
Subtract from SL due to ADT cleaning
If no rain and street is cleaned, then SL = SL * (1 - CE)
If rain event, then SL = 0
Save SLs and other parameters of day
CASE 5: Order RDC
If rain event, then SL = 0
Add to SL due to pavement wear
Add to SL due to tire wear
If spill, then add to SL
Add to SL due to dry track-on
Add to SL due to mud track-on
Add to SL due to mud wash-on
Add to SL due to sand/salt
Subtract from SL due to ADT cleaning
If no rain and street is cleaned, then SL = SL * (1 - CE)
Save SLs and other parameters of day
MRI-M\R9712-77.12 1 5
-------
Figure 3-1 (Continued)
CASE 6: Order RCD
If rain event, then SL = 0
Subtract from SL due to ADT cleaning
If no rain and street is cleaned, then SL = SL * (1 - CE)
If spill, then add to SL
Add to SL due to dry track-on
Add to SL due to mud track-on
Add to SL due to mud wash-on
Add to SL due to sand/salt
Save SLs and other parameters of day
end while
calculate statistics
do Report
end
MRI-MNR9712-77.12 1 6
-------
The first step of the daily program loop will be to generate a rainfall event and
temperatures. High and load temperatures will be generated, initially using a normal
distribution with an RSD of 20%. The average of these two temperatures will be used
to determine if snow/ice occurs on a specific day (< 32°F). The METEOR data base
contains monthly high and low temperatures. For simplicity, only the mid-season
month will be used, i.e., February for winter, May for spring, August for summer, and
November for fall.
Each road class will be simulated with the same meteorology, but with different
ADTs, deposition and cleaning events. Each major road class will have a default
1-lane width of 12 ft. For each day and road class, events will be randomly ordered
by cleaning, deposition, and rainfall, giving six (3!) paths through the inner program
loop. After each day of simulated events, the parameters of the day will be saved,
including the sequence of activities and the silt loadings due to each event.
3.4 PROGRAM SEQUENCE
For simplicity, deposition (D) events will always be made to occur in the
following sequence:
1) Pavement wear
2) Tire wear
3) Truck load spills
4) Dry track-on of dirt
5) Mud track-on
5) Mud wash-on
6) Sanding/Salting
Pavement wear and tire wear will occur for each day in the simulation and will be
totally dependent on ADT. Truck load spills will occur as sporadic events dependent
on fraction of ADT for heavy trucks, trucks carrying loads, and fraction of trucks that
spill their loads. Track-on of dry dirt will be dependent on ADT; a second addition of
tracked-on mud will occur as a function of rainfall on the day of interest and on rainfall
of the previous three days. Both track-ons will occur only on local, collector, and
arterial roads with no curbs. Mud wash-on will be generated as a function of rainfall
MRI-MR9712-77.12 1 7
-------
quantity on the day of interest, and will be generated only on collector and local roads
with no curbs.
Cleaning (C) events will always be made to occur in the following sequence:
1) ADT
3) Street Sweeping
Street sweeping will not occur on days with rainfall. Rainfall (R) will be considered a
cleaning event as described below.
Path 1: Order CDR (Cleaning, Deposition, Rainfall)
This path will assume that all cleaning events occur before all deposition
events, and that rainfall, if occurring, will happen last and will leave the street clean at
the end of the day. If rainfall occurs on this day, the silt loading that should be saved
should be one-half the silt loading of the day after both cleaning and deposition events
have been considered. This is to simulate an average silt loading on a day with rain-
fall. Otherwise, the silt loading on the road should be the total silt loading after all
deposition events have occurred. It is expected that on days with no rainfall, these
days will exhibit higher than expected silt loadings due to the sequence of activities.
Path 2: Order OCR (Deposition, Cleaning, Rainfall)
This path assumes that all deposition events precede cleaning events. Rainfall,
if occurring, will happen last and will leave the street clean at the end of the day.
However, the silt loading that should be saved is one-half the silt loading at the end of
the day after all cleaning and deposition events have occurred. Otherwise if no rain
occurs, the silt loading should be the total silt loading after all cleaning events have
occurred. This sequence is expected to produce lower than expected silt loadings on
days with no rainfall.
MHI-MVH9712-77.12 1 8
-------
Path 3: Order CRD (Cleaning, Rainfall, Deposition)
This path assumes that all deposition events will occur after rainfall has
occurred. The effect will be to enhance silt loading because rainfall will increase
track-on and carry-on.
Path 4: Order DRC (Deposition, Rainfall, Cleaning)
This path does not allow the daily rainfall to influence deposition events, and
the effect will be to diminish increased silt loading due to track-on and wash-on.
Street sweepers will not operate on days with rainfall. However, silt loading will be
much less than expected due to cleaning of the road first by rainfall, then by ADT.
Path 5: Order RDC (Rainfall, Deposition, Cleaning)
In this path, daily rainfall will influence both deposition and cleaning events. A
street sweeper will never operate on a day with rainfall. However, silt loading is
expected to be somewhat less than expected due to the cleaning of the road by all the
day's traffic after all deposition events have occurred.
Path 6: Order RCD (Rainfall, Cleaning, Deposition)
Daily rainfall will influence both deposition and cleaning events. A street
sweeper will never operate on a day with rainfall. Silt loading is expected to be much
higher than expected because all deposition events will occur after all cleaning events,
and all deposition events will be influenced by rainfall, if it occurs.
Approximately 1,000 days in one season will be recommended to be simulated
before the basic program loop is exited. A fixed-field length formatted output file will
be created with a daily event log and the average silt loading for each day. A
statistical summary of the results of the simulation of the output will be created from
this file, but may also be analyzed using SAS®, 1-2-3®, FoxPro®, etc.
MRI-MVR9712-77.12 1 9
-------
SECTION 4
DATA STRUCTURES
4.1 PROGRAM DATA STRUCTURES
Table 4-1 is a listing from a data base of data bases associated with the
planned program to model street surface silt loadings. This list will serve as the
project data dictionary and identifies five other data bases that will be used by the
model, including RDCLASS, GENPARAM, METEOR, SIM_PARM, and RESULTS.
The RESULTS data base will contain the output from the simulation.
Each record in the RDCLASS data base will specify a 1-mi road segment to be
modeled. A basic four-record data base will include default information on interstates,
arterials, collectors, and local roads, and this data base may be expanded by the user
to further define road segments within a particular county or city. Field names within
the record are located in the second column of Table 4-1. The class name, the
location of the road, the ADT, whether asphalt or concrete, and whether curbed or
uncurbed are also identified for each road class.
The GENPARAM data base contains information on the city to be modeled,
especially the percentages of land use. This simplification allows percentages of
residential, commercial, and industrial uses to apply to all road classes in a
metropolitan area.
METEOR is the meteorology data base that can contribute date for cities
across the nation. This data base is modeled on one that was developed for the
TANKS program created by Perrin-Quarles for MRI. Statistical parameters associated
with rainfall amounts during four seasons will be added to this data base, but will be
generated only for several cities. The temperature data, accompanied by rainfall
occurrence, will be used to model snow events.
MRI-MVR9712-77.12 21
-------
TABLE 4-1. PARM DATA FORMATS
DB_NAME PIELD_NAME
FIELD SIZ FIELD TYPE
EX FLD NAM
DESCRIPT
UNITS
RDCLASS
RDCLASS
RDCLASS
RDCLASS
RDCLASS
RDCLASS
RDCLASS
OENPARAM
OENPARAM
OENPARAM
OENPARAM
OENPARAM
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
RD_CLAS_ID
CLASS_NAME
ROAD_CLASS
MACADAM
AVO_DLY_TR
LANES
CURB
LOCATION
NEAR_CITY
COMM_PCT
IND_PCT
RES_PCT
FALL_P11
FALL_P12
FALL_P21
FALL_P22
WINTER_P11
W1NTER_P12
WINTER_P21
WINTER_P22
SPRINO_P11
SPRINO_P12
SPRINO_P21
SPRINO_P22
SUMMER_P11
SUMMER_P12
SUMMBR_P21
SUMMER_P22
CITY
STATE
ABBREV
A_TEMP_DAA
TAX_JAN
TAX_FEB
TAX_MAR
TAX_APR
TAX_MAY
TAX_JUN
TAX_JUL
TAX_AUO
TAX_SEP
TAX_OCT
Road Class Id Number
Road Class Name
Road Class Type
Macadam / Concrete
Average Daily Traffic
Lanes of traffic
Curb
Location
Nearest city
Commercial percent
Industrial percent
Residential parcant
Fall Pll percentage
Fall P12 percentage
Fall P21 percentage
Fall P22 percentage
Winter Pll parcantaga
winter P12 parcantaga
Winter P21 parcantaga
Winter P22 parcantaga
Spring Pll parcantaga
Spring P12 parcantaga
Spring P21 parcantaga
Spring P22 parcantaga
Summer Pll parcantaga
Summer P12 parcantaga
Summer P21 parcantaga
Summer P22 pareantaga
City name
Stata nama
Stata abbreviation
Automatically assigned a sequential id # by program
Chosen by the system user
One of four types (interstate,arterial, collector, local)
True if surface is macadam, false if concrete
No of vehicle passes per day
Number of lanes (1-16)
True if curbed, false otherwise
City/Region
Choose nearest major city for meteorological data
percentage of land used commercially
percentage of land used industrially
percentage of land used residentially
Row 1 Col 1 of Rainfall matrix
Row 1 Col 2 of Rainfall matrix
1 of Rainfall matrix
2 of Rainfall matrix
1 of Rainfall matrix
2 of Rainfall matrix
Row 2 Col 1 of Rainfall matrix
Row 2 Col 2 of Rainfall matrix
Row 1 Col 1 of Rainfall matrix
Row l col 2 of Rainfall matrix
1 of Rainfall matrix
2 of Rainfall matrix
1 of Rainfall matrix
2 of Rainfall matrix
1 of Rainfall matrix
Row 2 Col 2 of Rainfall matrix
City where meteorological station is located
State where meteorological station is located
VMT
Row 2 Col
Row 2 Col
Row 1 Col
1
ROW
Col
Row 2 Col
Row 2 Col
Row 1 Col
Row 1 Col
Row 2 Col
Avg maximum temperature Average maximum temperature for each month
10 Character
10 Character
10 Character
1 Logical
6 Numeric
2 Numeric
1 Logical
30 character
20 character
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
4 Numeric
20 character
15 Character
2 character
6 Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
S Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
MRI-M\R9712-77.12
-------
TABLE 4-1 (Continued)
DB_NAME FIELD_NAME
FIELD SIZ FIELD TYPE
EX_FLD_NAM
DESCRIPT
UNITS
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
|\j METEOR
CO METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
METEOR
SILTPRAM
SILTPRAH
SILTPRAM
SILTPRAM
SILTPRAM
SILTPRAM
SILTPRAM
SILTPRAM
RESULTS
RESULTS
RESULTS
RESULTS
TAX_NOV
TAX_DEC
TAX_ANN
TAN_JAN
TAN_FEB
TAH_MAR
TAM_APR
TAN_MAY
TAN_JUN
TAN_JBL
TAH_AUO
TAM_SEP
TAN_OCT
TAN_NOV
TAN_DEC
TAH_ANK
INSOL.JAN
INSOL_FEB
INSOL_MAR
IHSOL_APR
INSOL_MAY
INSOL_JUN
INSOL_JUL
INSOL_AU(9
IHSOL_SEP
INSOL_OCT
INSOL_NOV
INSOL_DEC
INSOL_ANN
MIND
COMPLETE
RD_CLS_ID
SEASON
PARAH_NAME
DEPOSITION
DISTRIBUT
MEAN_ALPHA
STD_BETA
ROAD_CLASS
RD_CLS_ID
DAY
PARAM_NAME
SILT_AMNT
Annual avg max temperature yearly average maximum temperature
Avg minimum temperature Average minimum temperature for each month
Annual avg mln temperature yearly average minimum temperature
Solar Insolation
Road class id
Season
Parameter Name
Deposition/Removal
Distribution
Mean / Alpha
standard deviation /
Road Class Type
Road class id number
Day
Silt parameter name
Silt amount
Link to RDCLASS, may contain "Profile* for default values
Season for this set of distribution parameters
Name of this silting event (Cleaning, Spill, etc)
True if this represents a deposition, false if a removal
Type of distribution (normal, lognormal, poisson, gamma)
contains mean or alpha value, depending on distribution
beta contains either standard deviation or beta
One of four types (interstate, arterial, collector, local)
assigned id 8
Current day in the simulation (1..total number of days)
Identifies type of silting event (spill, cleaning, etc).
Amount of silting for this parameter (+deposit, -removal)
Depends
Depends
6 Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
£ Numeric
6 Numeric
6 Numeric
6 Numeric
£ Numeric
£ Numeric
6 Numeric
6 Numeric
fi Numeric
6 Numeric
fi Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
6 Numeric
fi Numeric
6 Numeric
£ Numeric
£ Numeric
£ Numeric
£ Numeric
£ Numeric
£ Numeric
1 Logical
10 Character
£ Character
character
Logical
Character
10 Numeric
10 Numeric
10 Character
10 character
5 Numeric
10 character
10 Numeric
10
1
10
MRI-M\R9712-77.12
-------
The data base, SIM_PARM, will contain data associated with deposition and
removal of silt loading for a particular road class and event type. For a deposition
event such as track-on during a particular season for a particular road, the parameters
of a statistical distribution will be given, including the mean, or alpha, value; and the
standard deviation, or beta, value. The model will use these statistical parameters to
generate a silt amount deposited on or removed from a road segment due to a
specific event.
All of the parameters stored in the input data bases described above can be
easily changed (via dBase, FoxPro®, or a utility within the program itself). The
algorithms, including those used for temporal decay of silt loading on a road, will be
built into the modeling program and will not be able to be changed by the user.
4.2 RESULTS DATA BASE
The planned output file of a simulation run will have the structure defined in
Figure 4-1. The advantage to the use of an xBASE or fixed-field length structure for
card type "5" is that statistical analysis can be performed on output data beyond that
reported in the program printout of results.
4.3 PROGRAM OUTPUT
An anticipated report format is shown in Figure 4-2 and will summarize the
simulation run for each road class. One page will be generated for each road class
and season to be simulated.
4.4 AUXILIARY DATA BASES
The silt loading data base presented in Appendix C is available for calibration
and testing purposes. This data base of measured silt loadings has been compiled by
MRI from MRI files, state air quality agencies, and other contractor reports. Street silt
loadings have been identified according to location, date of sampling, average daily
traffic (ADT), and reported road classification. Currently, MRI is engaged in collecting
silt samples from paved road surfaces at 5 sites in Kansas City and 2 sites in Denver.
This 1-year effort has been underway for about 3 months and should substantially
increase the data base of measured silt loading values.
MRI-M\R9712-77.12 24
-------
Figure 4-1. OUTPUT FILE STRUCTURE.
1 Date of run:
Time of run:
User:
2 City:
State:
Commercial %:
Industrial %:
Residential %:
Season:
Average maximum temperature:
Average minimum temperature:
Temperature distribution RSD:
Rainfall, Pin
Rainfall, P12
Rainfall, P21
Rainfall, P22
Rainfall, alpha
Rainfall, beta
3xxxxx Contents of RDCLASS
4xxxxx Contents of SILTPRAM
Sxxxxx Day No.
Rainfall amount (in)
Residual effect of previous rainfall (in)
High temperature (°F)
Low temperature (0F)
Road class ID
Daily ADT
Sequence (CDR, OCR, CRD, DRC, RDC, RCD)
Beginning SL of day (lb/1-lane mile)
+ SL due to pavement wear (lb/1 -lane mile)
+ SL due to tire wear
+ SL due to spills
+ SL due to dry track-on
+ SL due to mud track-on
+ SL due to mud wash-on
+ SL due to sand/salt
- SL due to ADT cleaning
- SL due to street sweeping
- SL due to rainfall
Ending SL of day
Average SL to date (g/m2)
MRI-M\R9712-77.12 25
-------
Figure 4-2. REPORT FORMAT (one page for each road class).
Date of run: City:
Time of run: State:
User: Season:
Output filename: (Allow user to name file, if saved)
Road class ID: Commercial (%):
Road class name: Industrial (%):
Curbs (Y/N): Residential (%):
No. of Lanes:
ADT:
Days with rainfall (%):
Days with sanding/salting (%):
Average ADT:
No. of days simulated:
Mean silt loading (g/m2):
Standard deviation, silt loading (g/m2):
Maximum daily silt loading (g/m2):
Minimum daily silt loading (g/m2):
SL Range (g/m2): Frequency
0.00 - 0.50
0.50 - 1.00
1.00 - 1.50
MRi-M\R9712-77.12 26
-------
SECTION 5
MODEL DEVELOPMENT
Flexibility and expansion capability will be designed into the model using the
previously described hardware, software environment and data structures. A proto-
type model will be initially developed to test the interface and the time required to run
1,000 days of simulated silt deposition and removal events.
Testing and calibration of the model will be limited to the use of existing data.
Validation of the model with independently generated data does not fall within the
scope of this project.
The deliverables (expected milestones) presented in Table 5-1 will be provided
to EPA during this work assignment. A final report will serve to document the
computer model to characterize street surface dust loading.
TABLE 5-1. SCHEDULE OF DELIVERABLES
Deliverable Planned date
Work Plan March 3, 1993
Specifications Document April 19, 1993
Computer Model April 25, 1993
Final Report/User Documentation April 28, 1993
The MRI Project Leader for this project is Mrs. Mary Ann Grelinger. As Project
Leader, Mrs. Grelinger will be responsible for day-to-day management of the project,
directing project staff, adhering to the project budget and schedule, and routine
MRI-MVR9712-77.12 27
-------
communications with the EPA Work Assignment Manager. Mrs. Grelinger will be
assisted primarily by Dr. Chatten Cowherd, Dr. John Cigas, and Mr. Lan Tran.
Development of program algorithms will be the responsibility of Dr. Chatten
Cowherd, Mrs. Mary Ann Grelinger, and Dr. Greg Muleski. Program modules will be
coded by Dr. John Cigas and Mr. Lan Tran. Dr. John Cigas will develop the program
interface, the data structures and the output modules. Mr. Tran will code the
algorithms. Dr. Jerry Flora will assess the ability of FoxPro® routines to generate
pseudo-random variates according to specified statistical distributions. Calibration and
testing of the system with existing data will be carried out by Mr. Tran and Mrs.
Grelinger. Mr. Michael Fischer will be responsible for assembling and writing
documentation of the system.
The MRI Program Manager for the Air Quality Management Division Contract,
Mr. Richard Crume, will provide project guidance and oversight. This responsibility will
include the review and approval of all major deliverables.
MRI-MVR9712-77.12 28
-------
SECTION 6
REFERENCES
1. Kinsey, J. S., S. Schliesser, and P.J. Englehart. Control Sources of PM10.
Draft Report, EPA Contract No. 68-02-3891, Midwest Research Institute,
Kansas City, Missouri, September 1985.
2. Cowherd, C., G. E. Muleski, and J. S. Kinsey. Control of Open Fugitive Dust
Sources. Final Report, EPA Contract No. 68-02-4395, Midwest Research
Institute, Kansas City, Missouri, September 1988.
3. Kinsey, J. S., Enhanced PM10 Methodology for Paved Roads. Volume I: Draft
Test Plan, EPA Contract No. 68-DO-0137, Midwest Research Institute, Kansas
City, Missouri, October 1992.
4. Kinsey, J. S., et al. Guidance Document for Selecting Antiskid Materials
Applied to Ice- and Snow-Covered Roadways. Final Report, EPA Contract
No. 68-DO-0123, Midwest Research Institute, Kansas City, Missouri, May 1991.
5. Kinsey, J. S., D. Hecht, and F. Pendleton. Inspection Manual for PM10
Emissions from Paved/Unpaved Roads and Storage Piles. Final Report, EPA
Contract No. 68-02-4463, Midwest Research Institute, Kansas City, Missouri,
October 1989.
6. State of California, Air Resources Board, Emission Inventory Branch, Technical
Support Division. Methods for Assessing Area Source Emissions in California.
September 1991.
7. Englehart, P. J., and G. E. Muleski. Open Fugitive Dust PM10 Control
Strategies Study. Final Report, South Coast Air Quality Management District,
MRI-M\H9712-77.12 29
-------
Contract No. 90059, Midwest Research Institute, Kansas City, Missouri, July
1990.
8. Englehart, P., and G. Muleski, Open Fugitive Dust PM10 Control Strategies
Study. Final Report, South Coast Air Quality Management District Contract
No. 90059, Midwest Research Institute, Kansas City, Missouri, October 1990.
9. Muleski, G., et al., Open Source PM10 Method Evaluation. Final Report, EPA
Contract No. 68-02-4463, Midwest Research Institute, Kansas City, Missouri,
March 1991.
10. Cowherd, C., C. M. Maxwell, and D. W. Nelson. Quantification of Dust
Entrainment from Paved Roadways. Final Report, EPA Contract
No. 68-02-1403, Midwest Research Institute, Kansas City, Missouri, July 1977.
11. Pendleton, F., et al., Street Sweeper Specifications. Draft Final Report,
Midwest Research Institute, Kansas City, Missouri, October 1990.
12. Kinsey, J. S., G. E. Muleski, M. A. Grelinger, and C. Cowherd. Urban Fugitive
Dust Test Needs. Final Report, EPA Contract No. 68-DO-0137, Midwest
Research Institute, Kansas City, Missouri, September 1991.
13. Emission Factor Development. Communication received from John Kinsey.
14. Road Sweeping Study. Received from John Kinsey.
MRI-MNH9712-77.12 30
-------
APPENDIX A
INPUT SCREEN DISPLAYS
MRI-MVR9712-77.12 A~1
-------
Add/Edit/Delete Road Class:
Input Road Class Information
Road Class Name:
Road Class:
Type of Surface:
Average Daily Traffic:
Number of Lanes:
« Save to disk »
Arterial
(•) Macadam
( ) Concrete
t ] Curbs
0
0
< Cancel
= Help==
Delete
Edit/Delete Road Class:
noose an existing road class=
class 2
rd class 3
rd class 4
test class
test 2
newest
= Kelt
MHI-M\H9712-77.12
-------
Edit Silting Event Profile:
=Edit Silting Parameters=
Silting Parameter:
Season:
Road Class:
^Cleaning
Spill
Fall
Arterial
Statistical Distribution:
Normal
( ) Deposition
(•) Removal
Mean (alpha): 45.0000
Standard Deviation (beta):
7.0000
« Save to disk » <
= Help=
Cancel
Edit General Parameter Profile:
=Edit General Road Class Parameters=
Location:- Escondido, CA
Nearest Major City:
San Antonio
'•San Diego
Land Use
Commercial: 30.0
Industrial: 30.0
Residential: 40.0
« Save to disk »
Cancel
Help=
MRI-MNR9712-77.12
-------
APPENDIX B
ANALYSIS OF PSEUDO-RANDOM NUMBERS
GENERATED USING FOXPRO ROUTINES
MRI-MVR9712-77.12 B"1
-------
INTEROFFICE COMMUNICATION
MIDWEST RESEARCH INSTITUTE
April 2, 1993
To: Mary Ann Grelinger
From: J.D.Flora 4& J'^^—
Subject: Project 9712-77-12 Goodness of Fit Tests for Pseudo-Random Numbers
The FOXPRO data base program is being considered for use as part of a
simulation study. Before this is undertaken, a check of the pseudo-random number
generator incorporated in that package was undertaken. Three pairs of sets of
approximately 2000 pseudo-random numbers were generated and subjected to
distributional tests. In each set of numbers, about 2000 uniform random numbers and
2000 normal (mean 100, standard deviation 15) random numbers were generated.
(Actually, there were 1999 numbers of each type in the first set, and 3999 generated
for the second two sets. It appears that the generation stops one number earlier than
expected, so in applications this should be adjusted for to provide the desired total.)
Several distributional tests were applied to the sets of pseudo-random numbers.
The sets were divided into 10 groups with equal probabilities and also into 20 groups
with equal probabilities. The frequency of each such group was tabulated and used to
construct a chi-squared goodness of fit test. The Kolomogorov-Smirnov test was used
to check the fit of the cumulative distribution to the hypothesized distribution (the
Lillefors version of the Kolomogorov-Smirnov test was used for the normal data). Each
data set was divided at the theoretical mean and a runs test above and below the mean
was done as one check on randomness. Also, auto-correlations for up to 100 lags were
calculated to test for possible serial correlations. None of the auto-correlations was
significant; almost all were less than their standard errors. Results of the goodness of
fit tests are tabulated in Table 1. The data set, test used, value of the computed test
statistic, and significance level (P-Value) are tabulated. Small P-Values would lead to
rejection of the hypothesis that the pseudo-random numbers fit the hypothesized
distribution (or are independent).
The sample sizes were 1999 for the normal and uniform sets 1 and 2000 for all
other sets. (Only 3999 data were included in the second group; I added one random
variable of each type to give a full 2000). With this large a sample small differences
from the hypothesized distribution or from independence would be significant. This can
be seen in the relatively small changes in the value of some test statistics that led to
large differences in the significance levels.
None of the tests was significant at the 1 % level. A few of the tests were
significant at the 5 or 10% levels, which would be expected from random fluctuations.
There was some suggestion that the normal variables did not fit the tails of the normal
-------
distribution as well as might be desired. Most of the significant results appears to be
attributable to chance. The second and third sets of pseudo-random numbers were
generated to see if the marginally significant tests observed in the first set were part of
a pattern or simply resulted from chance. Comparison of the tests among the three
groups suggests that only chance is operating.
The overall conclusion is that the pseudo-random number generator seems to
meet the distribution tests satisfactorily so that it can be used in the planned simulation
experiment.
-------
TABLE 1: Goodness of Fit Test Results
Data Set
Normal 1
Normal 2
Normal 3
Normal 1
Normal 2
Normal 3
Normal 1
Normal 2
Normal 3
Normal 1
Normal 2
Normal 3
Uniform 1
Uniform 2
Uniform 3
Uniform 1
Uniform 2
Uniform 3
Uniform 1
Uniform 2
Uniform 3
Uniform 1
Uniform 2
Uniform 3
Test
Lillefors (K-S)
Lillefors (K-S)
Lillefors (K-S)
Chi-Squared 10
Chi-Squared 10
Chi-Squared 10
Chi-Squared 20
Chi-Squared 20
Chi-Squared 20
Runs
Runs
Runs
K-S
K-S
K-S
Chi-Squared 10
Chi-Squared 10
Chi-Squared 10
Chi-Squared 20
Chi-Squared 20
Chi-Squared 20
Runs
Runs
Runs
Statistic
0.0211
0.0142
0.0153
14.41
12.50
7.66
35.41
15.92
18.14
0.562
1.908
0.926
0.0267
0.0355
0.0144
13.75
14.94
10.25
27.66
33.32
17.02
1.155
-1.466
1.611
P-Value
0.0379
0.4153
0.314
0.107
0.187
0.569
0.0125
0.663
0.513
0.574
0.0564
0.355
0.115
0.0223
0.799
0.132
0.093
0.331
0.090
0.022
0.588
0.248
0.143
0.107
-------
250
200
150
100
50
0
0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95
Se.t 1 ID Cells
-------
Uniform 20 Cells
140
120
100
u 80
0)
cr
? 60
u.
40
20
0
Set 1 - 20 Cells
-------
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0
0.1 0.2 0.3 0.4
0.5
0.6 0.7 0.8 0.9
Sample
Uniform
U n'. f
a r wt
Set L
-------
1
.9
.8
,7
.6
.5
.4
.3
.2
.1
0
Uniform N=1999
0
>.0
1°
"o 0
>
1 °
1°
o 0
0
0.1 0.2 0.3
0.4 0.5 0.6
Value of X
0.7 0.8 0.9
Sample
Lower 90%
Upper 90%
L/l V\\ f A V
w\
-------
1
.9
.8
.7
.6
.5
,4
,3
,2
.1
0
Uniform N=1999
0
>-0
1 0
o
3o
S°
I °
1°
o 0
0
0
0.1 0.2 0.3
0.4 0.5 0.6
Value of X
0.7 0.8 0.9
Sample
Uniform
Lower 90%
Upper 90%
U v\l £ ovw\
Sd.L 1
-------
LJLJ
4
LU
0
K. -1
x
LLJ
-2
-3
-4
-4 -3 -2 -1
0
4
STDNORM
Set i
-------
Sample Normal CDF N=1999
1
0.9
0.8
0.7
0.6
y/r
ff
D
o
0.1
O-l r
-4 -3 -2-10 1
Value of X
Sample CDF Lower 95% Upper 95%
N/a ^ t^ a L S £. t 1
-------
250
200
150
100
50
-3.6441.06-0.68250.390.127512750.390.6829.0613.641
Set
1-10
C
-------
140
120
100
80
60
40
20
a®£ffiiffl9'&8 2915993 35
Set i - 10
-------
Second Data Uniform
260
240
220
200
180
160
140
120
100
60
60
40
20
0
I
\
I
:1
:sfc
:^
:$8
:^
:^
:^
:^
•^
:^
;^
;^
:^
:^
:^
:^
:^
:^
:^
"^
:^
> Ov>
m
0.1
0.2 0.3 0.4 0.5 0.6 0.7 0.6 03
Sample 2
Sample 3
lortoi
2 ahJ> 3 ~ Ib
-------
Second Data Uniform
130
120
no
100
90
60
70
60
50
40
30
20
10
0
0.05 0.15 025 0.35 0.45 0.55 0.65 0.75 0.85 0.95
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.6 0.9 1
Sample 2
Sample 3
2
~ lo
-------
I
£
240
220
200
180
160
140
120
100
60
60
40
20
0
Normal Two Samples of 2000
10 Group*
I
80.7745
87.373
92.134
962005
107.866
112.627
1192255
Sample 1
Sample 2
Sets 2 dkJ 3- 10
-------
130
120
no
100
90
60
70
60
50
40
30
20
10
0
Normal Two Samples of 2000
20 Groups
75.325 84.452569.8625 94.219 98.134 101.866 105.781 110.1175 (115.54751124.675
80.7745 87.373 92.134 96.2005 100 103.7995107.866 112.627 1192255
Sample 1
Sample 2
Sets 2 dvJ 3 - 20 Cells
-------
APPENDIX C
SILT LOADING DATA BASE
(From WA No. 44, EPA Contract No. 68-DO-0123)
MRI-M\R9712-77.12 C" 1
-------
• * denotes missing Information
STATE
CITY
STREET
CUSS
DATE
The following data from Reference 1
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
Billings
Billings
Missoula
Butte
Butte
Billings
Billings
Missoula
Butte
Missoula
Butte
East Helena
East Helena
East Helena
East Helena
Libby
Ubby
Libby
Libby
Ubby
Libby
Libby
Butte
Butte
Butte
Butte
Butte
East Helena
East Helena
East Helena
Columbia Falls
Columbia Falls
Columbia Falls
Columbia Falls
Columbia Falls
Columbia Falls
Columbia Falls
•Columbia Falls
Columbia Falls
Columbia Falls
Columbia Falls
East Helena
East Helena
East Helena
—
Yellowstone
Bancroft
1stSt
N Park PI
Grand Ave
4th Ave E
6th St
Harrison
Hiway 93
Montana
Thurman
1stSt
Montana
Main St
6th
5th
Champion Int So gate
Mineral Ave
Main Ave btwn 6th &
California
US 2
Garfield Ave
Continental Dr
Garfield Ave
So Park Ave
Continental Or
Morton St
Main St
US 12
7th St
4th St
3rd Ave
4th Ave
CF Forest
12th Ave
3rdSt
Nucleus
Plum Creek
6th Ave
US 2
Morton
Main St
US 12
Rural
Residential
Residential
Residential
Residential
Collector
Collector
Collector
Arterial
Arterial
Arterial
Residential
Local
Collector
Collector
Local
Local
Collector
Collector
Collector
Collector
Arterial
Residential
Arterial
Residential
Residential
Arterial
Local
Collector
Arterial
Residential
Residential
Residential
Residential
Local
Collector
Collector
Collector
Collector
Collector
Arterial
Residential
Collector
Arterial
04/78
04/78
04/78
04/78
04/78
04/78
04/78
04/78
04/78
04/78
04/78
04/83
04/83
04/83
04/83
03/88
03/88
03/88
03/88
03/88
03/88
03/88
04/88
04/88
06/89
06/89
06/89
08/89
08/89
08/89
03/90
03/90
03/90
03/90
03/90
03/90
03/90
03/90
03/90
03/90
03/90
07/90
07/90
07/90
SILT LOADING
DT
50
115
4000
679
60
6453
3328
3655
22849
18870
13529
140
780
2700
1360
1310
331
800
5900
536
4500
10850
562
5272
562
60
5272
250
2316
7900
390
400
50
1720
240
1510
1945
4730
316
1764
13110
250
2316
7900
(g/m*m)
0.6
0.5
8.4
24.6
103.7
1.6
7.7
26
1.9
1.9
0.8
13.1
4
8.2
4.7
7
61
2.1
0.9
1
2.8
7.2
1.7
0.7
2.1
18.8
_—
15.4
.__
__
2.7
1.6
5.6
3.2
TOTAL LOADINO
%
18.5
14.3
33.9
10.6
7
19.1
7.7
62.9
5
55.9
6.6
4.3
13.6
9.4
8.4
14.8
16.5
27.5
16
20.4
12.1
12.3
10.9
10.1
8.7
10.9
3.6
6.8
4.1
12.5
9.5
14.3
14.3
5.4
16.3
8.8
7
10
6.2
4.2
18.7
17
10.6
15.4
(g/m*m)
3.4
3.5
24.9
232.4
1480.8
13.05
99.5
6
37.3
3.3
11.9
305.2
29
86.6
55.3
—
43.5
299.2
19.3
8.8
11.2
25.5
197.6
24.6
17
16.5
131.5
—
— _
153.9
14.6
9.3
52.5
20.9
SILT LOADING SUMMARY
COMMENTS
2 samples, range: 1.0 - 2.2
-------
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT"
Columbia Falls
Ubby
Columbia Falls
Columbia Falls
East Helena
East Helena
East Helena
Columbia Falls
Columbia Falls
Libby
LJbbv
ULTUJ
Butte
East Helena
East Helena
East Helena
East Helena
East Helena
Thompson Falls
Thompson Falls
East Helena
East Helena
East Helena
Libby
UWJ
Libbv
WVJ
Butte
Butte
Kalispell
Kalispell
Thompson Falls
Thompson Falls
Helena
Kalispell
Columbia Falls
Kalispell
Thompson Falls
Thompson Falls
Libby
Libbv
i-iwisy
East Helena
East Helena
Thompson Falls
Thompson Falls
Libby
Libby
Kalispell
Columbia Falls
Kalispell
Columbia Falls
Columbia Falls
4th Ave
Main Ave 4th &
Nucleus
US 2
Morton
Main
US 12
Nucleus
US 2
US 2
Main Ave 4th &
Texas
King
Prickly Pear
Morton
MainSt
US 12
Preston
Highway 200
Seaver Park Rd
New Lake Helena Or
Potter
Main Ave 4th &
US 2
Texas
Harrison
3rd btwn Main & 1st
Main
Preston
Highway 200
Montana
3rd btwn Main & 1st
Nucleus
Main
Preston
Highway 200
Main Ave 4th &
US 2
Morton
US 12
Preston
Highway 200
Main Ave 4th &
US 2
3rd btwn Main & 1st
Nucleus
Main
Nucleus
US 2
Local
Collector
Collector
Arterial
Local
Collector
Arterial
Collector
Arterial
Arterial
Collector
Collector
Local
Local
Local
Collector
Arterial
Local
Collector
Local
Collector
Collector
Collector
Arterial
Collector
Arterial
Collector
Arterial
Local
Collector
Arterial
Collector
Collector
Arterial
Local
Collector
Collector
Arterial
Local
Arterial
Local
Collector
Collector
Arterial
Collector
Collector
Arterial
Collector
Arterial
08/90
08/90
08/90
08/90
10/90
10/90
10/90
11/06/90
11/06/90
12/08/90
12/09/90
12/13/90
01/91
01/91
01/91
01/91
01/91
01/23/91
01/23/91
02/91
02/91
02/91
02/14/91
02/17/91
02/21/91
02/21/91
02/24/91
02/24/91
02/25/91
02/25/91
03/91
03/09/91
03/91
03/09/91
03/91
03/91
03/91
03/91
04/91
04/91
04/91
04/04/91
04/91
04/91
04/14/91
04/91
04/14/91
05/91
05/91
400
530
5730
13039
250
2316
7900
5670
15890
10000
530
3070
75
425
250
2316
7900
920
5000
150
2140
850
530
10000
3070
22849
2653
14730
920
5000
21900
2653
5670
14730
920
5000
530
11963
250
7900
920
5000
530
12945
2653
5670
14730
5670
14712
1.5
2.4
0.8
0.2
3.4
4.5
0.6
5.2
1.7
21.5
13.6
1
1
12
14.1
36.7
0.8
9.2
33.3
21.6
19.2
74.4
33.3
69.3
1.2
2.9
30.5
17.4
35.7
66.8
15.4
39.1
30.1
17.6
4.4
4.3
14.8
20
4.3
0.5
1.2
2
3.5
11.8
15.1
9
13
2.4
5.5
4
17.9
5.3
7
10.2
5.6
13.9
13.5
24.1
9.6
27.1
15.4
3.4
1.8
3.5
12.1
14
9.9
27.2
7.1
9
7.7
18.7
21
11
7.9
24.8
20.4
17.9
17.8
6.2
29.1
17
24.7
8.3
15.5
33.1
19.5
8.8
8.7
15.7
13.4
44
20.5
37.1
19.8
44.5
• 17.5
20.7
37.7
13.2
16
o o
2.9
33.6
81.3
4.3
38
7.2
223.9
50.3
6.4
30.6
666.5
402.3
303.4
5.6
93
122.2
304.7
213.4
966.8
178.2
330.3
10.9
36.6
122.9
85.2
199.6
375.3
248.3
134.5
174.6
71.4
51
28.9
44.9
111.9
48.7
5.7
fi i
u.*3
14.7
7.8
57.2
40.9
47.6
29.4
15.9
24.8
2 samples, range: 0.8 - 0.8
2 samples, range: 2.8 - 5.9
2 samples, range: 1.0 - 7.5
2 samples, range: 13.5-16.1
3 samples, range: 11.4 - 32.4
4 samples, range: 0.3 - 4.0
2 samples, range: 1.1-2.2
2 samples, range: 2.5 - 4.4
4 samples, range: 1.2 - 22.9
4 samples, range: 1.3 - 3.8
5 samples, range: 1.5 -14.2
-------
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
Libby
Llbby
Ubby
East Helena
East Helena
Thompson Falls
Thompson Falls
Helena
Butte
Butte
Kalispell
Kalispell
Columbia Falls
Missoula
East Helena
Butte
Butte
Kalispell
Kalispell
East Helena
East Helena
East Helena
Columbia Falls
Columbia Falls
Kalispell
Kalispell
Thompson Falls
Thompson Falls
Butte
Butte
East Helena
Libby
Ubby
East Helena
Columbia Falls
Columbia Falls
East Helena
Thompson Falls
Thompson Falls
Kalispell
Kalispell
Kalispell
Thompson Falls
Thompson Falls
Kalispell
Kalispell
Kalispell
Kalispell
Kalispell
Main Ave 4th &
Main Ave 4th &
US 2
Morton
Main
Preston
Highway 200
Montana
Texas
Harrison
3rd btwn Main & 1st
Main
US 2
Russel btwn 4th & 5th
US 12
Texas
Harrison
3rd btwn Main & 1st
Main
Morton
Main St
US 12
Nucleus
US 2
3rd btwn Main & 1st
Main
Preston
Highway 200
Texas
Harrison
Morton
W 4th St
Main Ave 4th &
Main St
Nucleus
US 2
US 12
Preston
Highway 200
3rd btwn 2nd & 3rd
3rd btwn Main & 1st
Main
Preston
Highway 200
3rd btwn 2nd & 3rd
3rd btwn Main & 1st
Main
3rd btwn 2nd & 3rd
3rd btwn Main & 1st
Collector
Collector
Arterial
Local
Collector
Local
Collector
Arterial
Collector
Arterial
Collector
Arterial
Arterial
Road
Arterial
Collector
Arterial
Collector
Arterial
Local
Collector
Arterial
Collector
Arterial
Collector
Arterial
Local
Collector
Collector
Arterial
Local
Local
Collector
Collector
Collector
Arterial
Arterial
Local
Collector
Local
Collector
Arterial
Local
Collector
Local
Collector
Arterial
Local
Collector
05/19/91
06/27/91
06/27/91
07/91
07/91
07/09/91
07/09/91
07/17/91
07/26/91
07/26/91
08/03/91
08/03/91
08/11/91
08/30/91
08/30/91
10/03/91
10/03/91
10/06/91
10/06/91
10/16/91
10/16/91
10/16/91
10/20/91
10/20/91
11/06/91
11/28/91
12/17/91
12/17/91
02/02/92
02/02/92
02/03/92
02/03/92
02/03/92
02/03/92
02/03/92
02/92
02/03/92
02/22/92
02/22/92
03/15/92
03/15/92
03/15/92
04/92
04/92
04/26/92
04/26/92
04/26/92
05/92
05/92
530
530
10000
250
2316
920
5000
21900
3070
22849
2653
14730
15890
5270
7900
3070
22849
2653
14730
250
2316
7900
5670
15890
2653
14730
920
5000
3070
22849
250
350
530
2316
5670
12945
7900
920
5000
450
2653
14730
920
5000
450
2653
14730
450
2653
1.7
1.7
3.8
1.7
8.8
10.9
2.1
0.9
2.5
1.6
5.8
4
0.1
1.6
7
1
2.1
10
4.3
1.8
1.6
1
1.9
1.2
2.2
2.7
4
1.5
19.1
8.3
78.3
36.3
10.7
57.9
29.2
51.3
2.9
0.5
1.2
40.2
81.1
16.5
0.43
0.8
20.9
19.2
10.7
8.3
8.5
31
24.3
12.6
11.4
11
11
8.1
4.7
28.2
28.2
23
21
5.6
8.3
20.5
17.7
23.1
31.3
27.7
31
20.5
6.7
13.9
11.3
12.3
8.6
18.1
13.2
11.6
12
9.5
56.3
49.9
14.8
20.1
32.2
14.3
18
14.6
11.9
37.3
32.1
14.9
18.2
45.8
50.9
33.5
35.6
32.4
5.7
7.1
30.6
15.3
79.7
98.7
25.9
19.4
8.9
5.8
25.3
19.3
2.3
19.3
34.3
5.4
9.1
31.9
15.7
5.9
7.7
14.9
13.3
10.2
17.8
30.8
22.5
11.6
164.5
69.3
824.7
64.5
21.4
391
145.4
143.1
20.7
2.6
8.1
338
217.3
51.3
3.2
4.7
45.5
37.7
32.1
23.5
25.8
2 samples, range: 13.0 - 89.5
3 samples, range: 0.4 -1.0
3 samples, range: 6.6 -10.3
3 samples, range: 6.3 -11.4
-------
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
MT
Kalispell Main
Libby W 4th St
Libby Main Ave 4th &
Libby US 2
East Helena Morton
East Helena Main St
East Helena US 12
Columbia Falls Nucleus
Missoula Inez btwn 4th & 5th
Missoula Russel btwn 3rd & 4th
Missoula 3rd btwn Prince & In
Arterial
Local
Collector
Arterial
Local
Collector
Arterial
Collector
Local
Collector
Arterial
05/92
05/11/92
05/11/92
05/92
05/15/92
05/15/92
05/15/92
05/25/92
06/04/92
06/04/92
06/04/92
Th» following data from Reference 2 & 3
CO
CO
CO
CO
CO
CO
CO
CO
Denver E. Colfax
Denver E. Colfax
Denver YorkSt
Denver E. Belleview
Denver I-225
Denver W. Evans
Denver W. Evans
Denver E. Louisiana
Principal Arterial
Principal Arterial
Principal Arterial
Principal Arterial
Expressway
Principal Arterial
Principal Arterial
Minor Arterial
03/89
04/89
04/89
04/89
04/89
05/89
06/89
06/89
The following data from Reference 4 & 3
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
Denver E. Louisiana
Denver E. Jewell Ave
Denver State Highway 36
Denver State Highway 36
Denver W. Evans Ave
Denver E. Mexico St
Denver E. Colfax Ave
Denver State Highway 36
Denver E. Louisiana Ave
Denver W. Evans Ave
Denver W. Colfax Ave
Denver Parker Rd
Denver W. Byron PI
Denver E. Colfax Ave
Minor Arterial
Collector
Expressway
Expressway
Principal Arterial
Local
Principal Arterial
Expressway
Minor Arterial
Principal Arterial
Principal Arterial
Local
Principal Arterial
Principal Arterial
01/90
01/24/90
01/30/90
02/01/90
02/03/90
02/07/90
02/90
03/90
03/10/90
03/90
03/90
04/90
04/90
04/18/90
The following data from Reference 5 .
UT
UT
UT
UT
UT
Salt Lake County 700 East
Salt Lake County State St
Salt Lake County I-80
Salt Lake County 1-15
Salt Lake County 400 East
Arterial
Collector
Freeway
Freeway
Local
*
*
*
*
*
The following data from Reference 6
NV
NV
NV
NV
NV
NV
NV
Las Vegas Lake Mead
Las Vegas Perliter
Las Vegas Bruce
Las Vegas Stewart
Las Vegas Ambler
Las Vegas 28th St
Las Vegas Lake Mead
Major
Local
Collector
Major
Local
Collector
Major
07/15/87
07/15/87
07/15/87
09/29/87
09/29/87
09/29/87
10/07/87
14730
350
530
12945
250
2316
7900
5670
500
5270
12000
1994*
2228*
780*
—
4731 *
1905*
1655*
515*
5.1
13.4
5.6
10.4
6.9
6.4
1.2
1
1
15.2
2
0.21
0.73
0.86
0.07
0.02
0.76
0.71
0.14
1.44*
2.24*
0.56*
1.92*
1.64*
2.58*
0.09*
23.6
56.5
58.9
25.6
6.7
10.2
6.9
21.7
17.4
14
13.1
2
1.7
1.2
4.2
3.6
1.9
1.2
4.66
42340
27140
77040
146180
5000
1.27*
0.41 *
0.05*
0.3*
0.21 *
0.137
0.288
0.023
0.096
1.967
0.81
2.23
1.64
0.38
1.38
0.52
0.19
11.5
17
21.4
23.5
4.07
12.4
31.2
26.1
24
23
15.8
14.9
21.7
23.7
9.4
29.4
103
62.8
17
4.5
5.6
108.4
15.7
19.9
106.7
74.8
2
0.4
74
66.1
3.5
3 samples, range: 3.8 - 5.9
1.187
1.692
0.1
0.419
46.043
6.51
7.14
6.3
1.63
6.32
3.4
1.26
4 samples, range: 0.04 - 0.47
18 samples, range: 0.08 -1.76
2 samples, range: 0.83 - 0.89
3 samples, range: 0.03 - 0.09
3 samples, range: 0.01 - 0.02
11 samples, range: 0.03 - 2.24
12 samples, range: 0.07 - 3.34
5 samples, range: 0.08 - 0.24
6 samples, range: 0.12 - 2.8
2 samples, range: 0.56 - 0.56
4 samples, range: 1.92-1.92
2 samples, range: 1.64-1.64
3 samples, range: 2.58 - 2.58
16 samples, range: 0.02 - 0.17
7 samples
3 samples
5 samples, range: 0.07 - 3.38
21 samples, range: 0.04 - 2.61
6 samples, range: 0.01 - 0.11
6 samples, range: 0.21 - 0.35
4 samples, range: 0.107 - 0.162
4 samples, range: 0.212 - 0.357
5 samples, range: 0.011 - 0.034
6 samples, range: 0.078 - 0.126
14 samples, range: 0.177 - 5.772
3 samples, range: 0.24 - 0.46
3 samples, range: 0.64 - 2.00
3 samples, range: 0.51 - 0.54
2 samples, range: 0.17 - 0.20
* Traffic Volume Hourly-traffic
counted during testing only.
* Road CLASS taken from
reference 3
* Samples are said to be wet
seived.
* Road CLASS taken from
reference 3
* Samples given specified as
"post storm."
-------
NV
NV
The follow
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
AZ
The follow
KS
MO
MO
KS
KS
MO
MO
IL
IL
TnO ffOliOV
MN
MN
Las Vegas
Las Vegas
ring data from R«
Phoenix
Phoenix
Phoenix
Glendale
Glendale
Mesa
Mesa
Phoenix
Mesa
Phoenix
Phoenix
Phoenix
Plma County
Pima County
Plma County
Pima County
Pima County
Pima County
Pima County
Pima County
Pima County
ring data from Re
Kansas City
Kansas City
Kansas City
Tonganoxie
Kansas City
St Louis
SL Louis
GraniteCity
GraniteCity
ring data from Re
Duluth
Duluth
Pertfter
Bruce
ferenc* 7
Broadway
South Central
Indian School & 28th
43rd & Vista
59th & Peoria
Mesa Drive
E. McKellips & Olive
17th & Highland
3rd & Miller
Avatoni25th
Apache
N. 28th St & E. Qlenros
6th Ave
Speedway Blvd
22nd St
Amklam Rd
Fort Lowel Rd
Oracle Rd
InnRd
Orange Grove
La Canada
'erenc* 8
7th
Volker
Rockhill
4th
7th
1-44
Kingshighway
24th
Benton
'erence 9
US53north
US53south
Local
Collector
Arterial
Arterial
Arterial
Arterial
Arterial
Arterial
Arterial
Collector
Collector
Collector
Collector
Collector
Collector
Arterial
Arterial
Collector
Arterial
Arterial
Arterial
Arterial
Arterial
Arterial
Arterial
Arterial
Collector
Arterial
Expressway
Collector
Arterial
Collector
Highway
Highway
10/07/87
10/07/87
02/80
02/80
02/80
03/80
03/80
05/80
05/80
05/80
05/80
03/19/92
02/26/92
5000
5000
1.5
0.9
0.127
0.085
0.035
0.042
0.099
0.099
0.014
0.028
0.07
0.528
0.282
0.035
1.282
0.401
0.028
0.014
0.113
0.014
0.021
0.162
0.106
0.29
0.67
0.68
2.5
0.29
0.02
0.08
0.78
0.93
0.23
0.24
31.9
24.1
12.2
5
3.1
3.9
8.2
8.9
17
13.4
11.8
11.1
6.4
2.3
6.417
8.117
16.529
5.506
3.509
1.556
18.756
21.989
3.975
6.8
20.1
21.7
14.5
12.2
10.9
6.4
8.6
28
13.4
4.76
3.74
1.071
1.726
1.021
1.049
1.183
1.085
0.092
0.232
0.627
4.79
4.367
1.479
19.961
4.937
0.176
0.197
3.268
0.725
0.127
0.725
2.571
4.2
3.5
3.3
17.1
2.4
0.7
12.3
10.8
1.94
2.3
2 samples, range: 1.48 -1.52
2 samples, range: 0.76 -1.03
* Samples not given a specific
date.
3 samples, range: 0.15 - 0.46
3 samples, range: 0.43 -1.00
4 samples, range: 0.02
3 samples, range: 0.05 - 0.11
2 samples, range: 0.73 - 0.83
8 samples, range: 0.04 - 0.77
5 samples, range: 0.05 - 0.37
------- |