development of criteria to identify pedestrian high crash …trc.faculty.unlv.edu/ndotrpt3.pdf ·...
TRANSCRIPT
Development of Criteria to Identify Pedestrian
High Crash Locations in Nevada
Quarterly Progress Report
Submitted to
Nevada Department of Transportation (NDOT) Research Division
1263 South Stewart Street Carson City, NV 89712
Krishna Kumar K. Vanjeeswaran Srinivas S. Pulugurtha
Shashi S. Nambisan
Transportation Research Center Howard R. Hughes College of Engineering
University of Nevada, Las Vegas 4505 S. Maryland Parkway Box 454007
Las Vegas, NV 89154-4007 Telephone: (702) 895-1338
Fax: (702) 895-4401
December 30, 2003
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
1
Development of Criteria to Identify Pedestrian High Crash Locations in Nevada
Nevada has experienced over 40 pedestrian fatal crashes per year over the last six years.
Likewise, Nevada also has experienced over 800 pedestrian injury crashes per year
during the same period. More than 70 percent of these pedestrian fatal crashes and
pedestrian injury crashes are in Clark County, Nevada. There is a critical pedestrian
safety issue on many urban streets in Nevada, in general, and in the Las Vegas
metropolitan area in Clark County, Nevada, in particular. The Las Vegas metropolitan
area is ranked among the worst urban areas in terms of pedestrian safety. Crashes in such
environment also result in adverse publicity, which can linger long after the incidents
themselves. Besides the adverse publicity, these crashes results in severe health and
human life consequences, and monetary impacts.
The main objective of this research project is to develop criteria to identify
“pedestrian high crash location” in order to allocate recourses including Federal Safety
Funds, for safety improvements. The criteria will help in the development of a
“Pedestrian Safety Program”, as a part of Nevada Department of Transportation’s
(NDOT) Federal Highway Safety Improvement Program (HSIP). The developed criteria
will assist the system managers not only in Las Vegas and Nevada, but also nationally, in
better understanding the cause of the crashes and identifying appropriate operating
strategies to enhance pedestrian safety.
The proposed research is divided into the following main tasks:
1. Task 1: Literature Research
2. Task 2: Data Collection and Geocoding
3. Task 3: Analysis of Pedestrian Data
4. Task 4: Develop Criteria to Identify “High Crash Locations”
5. Task 5: Recommendations and Scope for Further Research
6. Task 6: Preparation of Progress Reports, Final Report and Publications
The focus during this quarter (October 2003 to December 2003) was primarily on
Task 4: Develop Criteria to Identify “High Crash Locations” which includes creating
Density Maps to identify locations having higher crash concentrations and, selecting
potential “High Risk Zones” with respect to the crash density.
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
2
Density Map
The geocoded pedestrian crashes show some clustering and some dispersion throughout
the study area. Several crashes occur at one single point, so the presence of a dot does not
necessarily equal one crash. For example, FIGURE 1 shows the spatial distributions of
pedestrian crashes in the City of Reno and the map does not exactly reflect the crash
concentrations of locations having more than one crash. Virginia Street and 4th Street
intersection in the map actually has 9 crashes, where as Virginia Street and Plaza Street
intersection has only one crash. But the map does not make any difference in representing
these crash concentrations. In order to identify the concentration and pattern of crashes,
which is important to locate pedestrian high crash locations, density map feature
available in ArcMap is used.
Density surfaces are used to demonstrate concentrations of point or line locations. For
example, if on an annual basis higher number of pedestrian crashes occurs on an
intersection than other locations, then the density of pedestrian crashes will be
concentrated near the intersection. Density is a measure of the quantity of something per
unit of area, such as the number of pedestrian crashes per square mile or people per
square mile. Density can be calculated using two methods: simple and kernel. A circular
search area is used by both methods to calculate density.
Simple Density
The simple method divides the entire study area to predetermined number of cells and
draws a circular neighborhood around each cell to calculate the individual cell density
values, which is the ratio of number of features that fall within the search area to the size
of the area (FIGURE 2). Radius of the circular neighborhood affects the resulting density
map. If the radius is more, higher the possibility that the circular neighborhood include
more feature points which results in a smoother density surface.
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
3
4th
5th
Plaza
Lake
Center
Virginia
Sierra
Commercial
Douglas
3rd
Plaza
¯0.05 0 0.050.025 Miles
LegendPedestrian Crashes
Street Network
FIGURE 1 Spatial Distribution of Pedestrian Crashes in the City of Reno (Zoomed in)
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
4
FIGURE 2 Simple Density Calculations (Source: ESRI VIRTUAL CAMPUS)
Kernel Density
Kernel method uses a mathematically complicated way to estimate the density compared
to the simple method. The kernel method divides the entire study area to predetermined
number of cells. Rather than considering a circular neighborhood around each cell
(simple method), kernel method draws a circular neighborhood around each feature point
and then a math function is applied that goes from 1 at the position of the feature point to
0 at the neighborhood boundary. Radius of the circular neighborhood affects the resulting
density map. If the radius is more, the flatter is the kernel. ArcMap 8.2 uses a Quatric
function to do the kernel density estimation. Density at a distance of r from sample point
= K * ( 1 – (rR
)2 ) 2 if r < R and 0 if r >= R
R = Search Radius
r = Distance from the sample point
K = 3
2π R
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
5
For example for a search radius of 500m the density can be calculated as (FIGURE 3):
Density at (r=0) i.e. at (0, 0) = 3
5002π (1 – (
0500
)2 ) 2
= 3.82 per sq. km.
FIGURE 3 Kernel Density Calculations
This kernel function is applied to each feature point and individual cell density values is
the sum of the overlapping kernel values over that cell divided by the area of the search
radius (FIGURE 4). A smoother looking density surface is created by kernel density
calculations than the simple density calculations.
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
6
FIGURE 4 Kernel Density: Calculating the Individual Cell Density Values (Source:
ESRI VIRTUAL CAMPUS)
For calculating the crash densities, kernel method is employed with a search radius of
400 feet. The resulting crash densities (corresponding to FIGURE 1) for the city of Reno
are shown in FIGURE 5. The map makes clear distinction between the crash
concentrations of locations having more number of crashes. From FIGURE 5 it is more
apparent that the Virginia Street and 4th Street intersection in the City of Reno has higher
crash concentrations compared to other intersections nearby.
Density Map is drawn for Washoe, Carson, Elko and Douglas Counties and crash
concentrations in each county is identified. The resulting crash densities are shown in
Figure 6 through Figure 9.
Development of Criteria to Identify High Pedestrian Crash Location Quarterly Progress Report
7
4th
5th
Plaza
Lake
Sierra
Center
Douglas
Virginia
Commercial
Evans
3rd
Plaza
¯0.025 0 0.0250.0125 Miles
LegendCrashes
Street Network
Crash Density0 - 229
229-458
458-687
687-916
916-1145
1145-1375
1375-1604
1604-1833
1833-2062
FIGURE 5 Pedestrian Crash Concentrations in the City of Reno (Zoomed In)
Development of Criteria to Identify High Pedestrian Crash Location Quarterly Progress Report
8
2nd
4thMill
PlumbVirginia
Arlington
Vassar
Plumas
California
Kuenzli
Monroe
Sutr o
¯0.2 0 0.20.1 Miles
LegendCrashes
Street Network
Crash Density0 - 229
229-458
458-687
687-916
916-1145
1145-1375
1375-1604
1604-1833
1833-2062
FIGURE 6 Pedestrian Crash Concentrations in the City of Reno
Development of Criteria to Identify High Pedestrian Crash Location Quarterly Progress Report
9
5th
US50
Roop
US 39 5
College
Robinson
Edmonds
Winnie
Salim
an
Cars
on R
iver
Lone
Mt n
¯0.25 0 0.250.125 Miles Legend
Crashes
Street Network
Crash Density0 - 179
179 - 359
359 - 538
538 - 718
718 - 897
897-1077
1077 - 1256
1256 - 1436
1436 - 1615
FIGURE 7 Pedestrian Crash Concentrations in the Carson City
Development of Criteria to Identify High Pedestrian Crash Location Quarterly Progress Report
10
5th
I-80
Idaho
Silver
9th
Cedar
2nd
12th
Wate
r
College
Copper
Argent
State Highway 225W
ilson
Douglas
6th
11th
Carlin
ColonialW
illow
Railroad
Douglas
Willow
6th11th
2ndC
olle
ge
11th
6th
College
6th
6th
Railroad
9th
¯0.25 0 0.250.125 MilesLegend
Crashes
Street Network
Crash Density0-79
79-158
158-237
237-316
316-395
395-474
474-553
553-632
632-711
FIGURE 8 Pedestrian Crash Concentrations in the City of Elko
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
11
Identifying the High Risk Zones
After identifying crash densities, the next step is to select potential “High Risk Zones”.
According to the FHWA Zone Guide for Pedestrian Safety (1998) zone process provides
a systematic method for targeting pedestrian safety improvements in a cost effective
manner. Zoning identifies a subset of jurisdiction containing as much of the pedestrian
problem of interest in as little land area as possible. Once zones are defined, pedestrian
safety programs can be focused in them with greatly increased efficiency. Initially the
methodology is applied to Washoe County, which is having 762 pedestrian crashes in
187.6 sq. miles of study area.
The map is examined for high pedestrian crash density that occurs along a single
strip of corridor. According to the FHWA Zone guide for an annual crash rate on order of
200, those roadway segments where six or more crashes occur in a two-mile segment
should be identified as linear zones. Thus for a study of 726 crashes, the minimum
number of crashes required to qualify as a linear zone is 22 crashes in a two-mile
segment. This crash rate is adjusted with respect to the segment length and 28 high risk
linear zones are identified for the study area (FIGURE 7).
For the locations having higher crash density which does not adhere along a
corridor is selected as individual circular zones with 300 feet radius and 31 similar high
risk circular zones are identified.
Finally for all zones combined, the percentages of both crashes and land area
covered are calculated in order to determine program coverage efficiency.
Ratio of percent of the problem addressed = Number of crashes inside all zones combined
Total number of crashes in the study area
= 460726
= 60.00 %
Ratio of the land area covered = Total Area of Linear Zones+Total area of Circular zones
Area of Study Area
= 0 629 0 313
187 6. . . .
. .sq miles sq miles
sq miles+
= 0.50 %
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
12
Efficiency ratio = Ratio of percent of the problem addressed
Ratio of the land area covered
= 6005.
= 120
Thus an efficiency ratio of 120 is obtained, which is much higher than the minimum
efficiency ratio of 3 specified by FHWA Zone Guide.
Similar methodology is applied to Carson, Elko, and Douglas to identify the “High Risk
Zones”. Five linear zones and five circular zones in Carson, 21 circular zones in Elko,
and 19 circular zones in Douglas are identified. TABLE 1 shows the liner high risk zones
selected in Washoe. TABLE 2 shows the circular high risk zones selected in Washoe.
TABLE 3 shows linear high risk zones selected in Carson. TABLE 4 shows circular high
risk zones selected in Carson. TABLE 5 shows circular high risk zones selected in Elko.
TABLE 6 shows circular high risk zones selected in Douglas. FIGURE 9 shows selected
high risk zones in Washoe. FIGURE 10 shows the selected high risk zones in Carson.
FIGURE 11 shows selected high risk zones in Elko. FIGURE 12 shows selected high risk
zones in Douglas.
The focus during the next quarter is to identify “High Risk Zones” in Clark and to
develop criteria to identify pedestrian “High Crash Locations”.
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
13
TABLE 1 Linear High Risk Zones Selected in Washoe
ZONE # ZONE1 4th Street: Lake Street to Keystone Avenue2 Virginia Street: 6th Street to 1st Street3 2nd Street: Lake Street to Keystone Place4 Arlington Avenue: 6th Street to Island Avenue5 California Street: Virginia Street to Hill Street6 Keystone Avenue: Sunnyside Drive to 5th Street7 Sierra Street: College Drive to 10th Street8 Virginia Street: College Drive to 9th Street9 Montello Street: Oliver Avenue to 9th Street
10 Sutro Street: Oliver Avenue to 4th Street11 Oddie Blvd: Sullivan Lane to Silverada Blvd12 El Rancho Drive: G Street to Prater Way13 Wells Avenue: Kuenzli Street to Mill Street14 Wells Avenue: Thoma Street to Taylor Street15 Mill Street: Kietzke Lane to Pringle Way16 Kirman Avenue: Mill Street to Ryland Street17 Rock Blvd: Glendale Avenue to Freeport Blvd18 Virginia Street: Pueblo Street to Plumb Lane19 Lakeside Drive: Plumb Lane to Hillcrest Drive20 Virginia Street: Linden Street to Peckham Lane21 Neil Road: Moana Lane to Peckham Lane22 Moana Lane: Kietzke Lane to Lakeside Drive23 Kietzke Lane: Plumb Lane to Gentry Way24 Grove Street: Wrondel Way to Kietzke Lane25 Brinkby Avenue: Robinhood Drive to Lakeside Drive26 Sun Valley Blvd: 7th Avenue to Scottsdale Road27 Baring Blvd: Springland Drive to Sparks Blvd28 Virginia Street: Bailey Drive to Talus Way
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
14
TABLE 2 Circular High Risk Zones Selected in Washoe
ZONE # ZONE1 Peckham Lane & Kietzke Lane2 Vassar Street & Kietzke Lane3 Vassar Street & Harvard Way4 Plumb Lane & Harvard Way5 Terminal Way & Mill Street6 Vassar Street & Locust Street7 Wells Avenue & Pueblo Street8 Stewart Street & Wells Avenue9 Second Street & Wells Street
10 Mill Street & Center Street11 7th & Center Street12 5th Street & Sierra Street13 Center Street & 5th Street14 Newland Circle & California Avenue15 7th Street & Elgin Avenue16 Silverada Blvd & Orchid Way17 9th Street & Shone Drive18 Prater Way & Sullivan Lane19 Plumb Lane & Arlington Avenue20 Sulivan Lane & Greenbrae Drive21 Tyler Way & Pyramid WAy22 Prater Way & I Street23 Pyramid Way & L Street24 Shadow Lane & Deep Creek Drive25 Greg Street & Sparks Blvd26 Stead Blvd & Silver Lake Road27 Colling Circle & Newport Lane28 2nd Street & Reservation Road29 Prosperity Street & Kietzke Lane30 Wells Avenue & 6th Street31 Plumb Lane & Locust Street
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
15
TABLE 3 Liner High Risk Zones Selected in Carson
ZONE # ZONE1 US395 Highway: Hotsprings Road to John Street2 US395 Highway: Caroline Street to 7th Street3 US 50 Highway: Stewart Street to Saliman Road4 US 50 Highway: Lompa Lane to Brown Street5 5th Street: Root Street to Saliman Road
TABLE 4 Circular High Risk Zones Selected in Carson
ZONE # ZONE1 Robinson Street & Saliman Road2 Winnie Lane & Lone Mtn Drive3 Hotsprings Road & Pine Lane4 College Parkway & US395 Highway5 US395 & Snyder Avenue
TABLE 5 Circular High Risk Zones Selected in Elko
ZONE # ZONE1 Cedar Street & 12th Street2 5th Street & Railroad Street3 Water Street & 6th Street4 9th Street & Douglas Street5 5th Street & Carlin Court6 Wilson Avenue & 6th Street7 Cedar Street & Buns Road8 Idaho Street & 11th Street9 Idaho Street & College Avenue
10 Silver Street & Elecart Blvd11 Idaho Street & Cedar Street12 Second Street & Willow Street13 Mittry Avenue & College Court14 Argent Avenue & Copper Street15 Antimony Road & Carlson Avenue16 Chris Avenue & Colonial Drive17 Spruce Road & Noodle Lane18 Kittridge Canyon Road & Lupine Street19 Spring Valley Pardway & Cedarlawn Drive20 Berry Creek Place & Berry Creek Drive21 Tres Cartes Avenue & Berry Creek Drive
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
16
TABLE 6 Circular High Risk Zones Selected in Douglas
ZONE # ZONE1 US 50 Highway & Elks Point Road2 US 50 Highway & Kingsbury Grade Road3 Tahoe Drive & Lynn Way4 Benjamin Drive & Tina Court5 Kingsbury Grade Road & Tramway Drive6 Tramay Drive & Jacks Circle (500 ft)7 Main Street & County Road8 Main Street & First Street9 Meadow Lane & Douglas Avenue
10 Main Street & Eddy Street11 US395 Highway & Kingslane Court12 Waterloo Lane & Toler Lane13 Muir Drive & Lyell Way14 Heritage Lane & Tillman Lane15 Main Street & Mill Street16 Mica Drive & Calcite Drive17 Tourmaline Drive & Granite Court18 Somerset Way & Plymouth Drive19 Sunridge Drive & Starshine Court
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
17
4th
Mill
Virginia
2nd
Plumb
Sutro
P lumas
VassarCalifornia
Kuenzli
Monroe
Arlington
4th
4th
2nd
I-80
McC
arra
n
Unit ed St at es H
igh way 395
¯0.9 0 0.90.45 Miles
LegendCrashes
Street Network
Circular Zones
Linear Zones
FIGURE 9 Selected High Risk Zones in Washoe
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
18
5th
US50
US395
Roo p
College
Edmonds
Salim
an
Winnie
Clearview
Fairview
Robinson
Carson River
Kings Canyon
Lone
Mtn
Clearview
¯0.5 0 0.50.25 Miles Legend
Crashes
Street Network
Linear Zones
Circular Zones
FIGURE 10 Selected High Risk Zones in Carson
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
19
5th
I-80
Idaho
Silver
9th
Cedar
2nd
12th
Wate
r
College
Copper
Argent
State Highway 225W
ilson
Douglas
6th
11th
Carlin
ColonialW
illow
Railroad
College
Col
lege
9th
Railroa
d
Water
6th
11th
Willow
Douglas
6th
6th
2nd
11th6th
¯0.3 0 0.30.15 MilesLegend
Crashes
Street Network
Circular Zones
FIGURE 11 Selected High Risk Zones in Elko
Development of Criteria to Identify Pedestrian High Crash Location Quarterly Progress Report
20
FoothillUS395
Stockyard
US5
0
Genoa
Muller
Jacks Valley
Stephanie
Mottsville
Kingsbury Grade
State High w
ay 88
E as t V a lle y
Old Kingsbury
US39 5
US50
Gen
oa
East Valley
LegendCrashes
Street Network
Circular Zones
2 0 21 Miles ¯
FIGURE 12 Selected High Risk Zones in Douglas