Evaluation of motorist warning systems for fog-related incidents in the Tampa Bay area

Evaluation of motorist warning systems for fog-related incidents in the Tampa Bay area

Material Information

Evaluation of motorist warning systems for fog-related incidents in the Tampa Bay area
Florida -- Dept. of Transportation
University of South Florida -- Center for Urban Transportation Research
Place of Publication:
University of South Florida, Center for Urban Transportation Research
Publication Date:
Physical Description:
46, [9] leaves : folded map, charts ; 28 cm.


Subjects / Keywords:
Highway communications -- Florida -- Tampa ( lcsh )
Fog -- Florida -- Tampa ( lcsh )
Weather forecasting -- Florida -- Tampa ( lcsh )
Traffic accidents -- Florida -- Tampa ( lcsh )
bibliography ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 43-45).
Additional Physical Form:
Also available online.
General Note:
In cooperation with the Florida Dept. of Transportation.
General Note:
"June 1997."
Statement of Responsibility:
prepared by Center for Urban Transportation Research, College of Engineering, University of South Florida.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
001931015 ( ALEPH )
37385386 ( OCLC )
C01-00302 ( USFLDC DOI )
c1.302 ( USFLDC Handle )

Postcard Information



This item has the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200289Ia 4500
controlfield tag 001 001931015
005 20080401120724.0
008 970730s1997 flubd b 000 0 eng d
datafield ind1 7 ind2 024
subfield code a C01-00302
2 local
b .E92 1997
0 245
Evaluation of motorist warning systems for fog-related incidents in the Tampa Bay area /
prepared by Center for Urban Transportation Research, College of Engineering, University of South Florida.
[Tampa] :
University of South Florida, Center for Urban Transportation Research,
46, [9] leaves :
folded map, charts ;
28 cm.
In cooperation with the Florida Dept. of Transportation.
"June 1997."
Includes bibliographical references (leaves 43-45).
Also available online.
Highway communications
z Florida
Weather forecasting
Traffic accidents
1 710
Dept. of Transportation.
University of South Florida.
Center for Urban Transportation Research.
8 773
t Center for Urban Transportation Research Publications [USF].
4 856
u http://digital.lib.usf.edu/?c1.302


Evaluation of Motorist Warning Systems for Fog-Related Incidents in the Tampa Bay Area prepared by Center for Urban Transportation R.esearcll College of Engineering, University of South Florida June 1997 The opinions, findings, and recommendations expressed in this report are those of the Center for Urban Transportation Research (CUT/() and the University of South Florida and not necessarily those of the Florida Department of Transportation. This report has been prepared in cooperation with the Florida Department ojTransponation-District VII Office, infuljillment of ENG No. ESC-DOT-96197-7007-TO, Journal Trans. 49-20-2-655006-489007-0050, and CUI'R Account No. 21-17-2 55-L. 0. CUTR Principal Investigator has been Micirdel C. Pietrzyk (ITS Program Manager), with assistance pr(Wided by Patricia A. Turner (Research Associate), Sandra L. Geahr (ITS Program Assistant) and Ramakrishna Apparaju (Graduate Research Assistant).


ACKNOWLEDGMENTS CUTR would like to thank a number of individuals who provided major contributions during the data collection portion of this evaluation. General meteorological background and input was provided by NewsChannel8 Chief Meteorologist David Grant, and Channell3 Meteorologist Howard Shapiro. Historical weather data and perspectives weie provided by Richard Rude and Anne Cornell at the National Weather Service in Ruskin, Florida; Bill Cheman with the Weather Squadron at MacDill Air Force Base ; and AI Chen at the National Climatic Data Center in Asheville, North Carolina. Crash report data were provided by Corporal Alan Hill, Hillsborough County Sheriff's Office ; Pat Lawrence, Hillsborough County Sheriff's Records Section; Dr. Richard Zeller, Florida Department of Highway Safety and Motor V chicles, Office of Management and Planning Services; an d Ray Boetch, Florida Department of Highway Safety and Motor Vehicles, Bureau of Office Services. Contacts for weather detection and incident warning systems were provided by Grant Zammit, Florida Division ITS Engineer, Federal Highway Administration. Princip3l contacts included Dr. Gary GimmestadGeorgia Tech Research Institute, Frank Simko ASTI Transportation Systems in New Jersey, Peter Allain Louisiana Department of Transportation & Development, David Cox-Federal Highway Administration Tennessee Division Office Don Dahlinger-Tennessee Department of Transportation, Stephen Ikerd-Federal Highway Administration South Carolina Division Office and Fred Kitchener-CH2M Hill. The assistance provided by all the aforementioned individuals has improved the quality of this report and is very much appreciated Principal contacts at the Florida Department of T1'811Sportation DiStrict Seven Office (Tampa) for this project were Keith Crawford David Buser, and John Temple.


TABLE OF CONTENTS I. Executive S11mmary .......... .......... .. ....... ................. .. . ... .................... ... .... ... ................ .... II BackgroUtl.d ............ ......... ....................... ....... .. ...... ... ............................................... ....... 2 III. Purpose ....... .. ...... .. ..... ............... ........................... ............................ . ........... .................. . .. 3 IV Meteorological Data Review ........... .... .... ......... ...... ..... ... ..... ........ ........................ . ......... ... 3 V. Crash Data Review . .. ............ .. ...... .... ..... ............ .. .. ...... .... .. ........ .. ... ........ .. ... .. ............. 9 VI Techn.o l ogies . .. ... ... .. ... .............. . ............ ... ..... .... .. .. .............................................. ... ... ..... 21 -Visibility Detection. ................ .............. ......... ....... ...... ........... . ..... ........... ............. ... 21 Incident Detection and Motorist Warning ..... . ......... ............ ...... ... ......... ..... ....... ....... 26 StlJDmazy .......... . .................... ...... ......... .... ..... ... ...... ... .................... ...... ...... .......... 33 VII. Driver Education andAwareness Techniques . . .. ... ... .. ...... ........ .......... ..... .. ..... ....... ......... 35 VIII. Conclusions and Recommendations ....... . ................... ... ................ .... ... ... ..................... . 41 Bibliography ..... ........................ ............ ...... .... ....... ............... .......................... ....... ... ..... 43 Appendix: Florida County Fog-Related Crash Rates by Y ear . ...... .......... ... .... .... ............. .46


LIST OF TABLES Paee Table 1 Selected Clima tological Data During "Fog Season" .. ........ ... ... ............... .......... .... .... ..... 7 Table 2-Motor V e hicle Crashes Fog-Related Crashes, and Crash Severity ........................ ...... 10 Table 3 -Most Common Driver/Pedestrian Contributing Causes in Fog-Related Crashes ......... .13 Table 4-Fog-Related Crashes by Time of Day .. ... ... ....... ................. : ....... .......... ......................... 14 Table 5 -Crash Injury Security in Fog-Related Crashes ...... .. ... ... .. ....... ........ ... ... ... .............. ........ .15 Table 6 Age Distribution of Drivers Involved in Fog-Rel ated Crashes ................ ..... ......... ...... .16 Table 7 Driver Residence Status in Fog-Related Crashes ....... ... .. ............ ............................... ... .! 6 Table 8-Type of Vehicle Involved in Fog-Related Crashes .................. ...... ........ ....................... 17 Table 9-Movement of Vehicle Involved in Fog-Related Crashes ... ...... ..... .. . .. .. ... .... ... ........ ..... 18 Table 10 -Roadway Type Where Fog-Related Crashes Occurred ...... .. ........ ........... ......... ....... .19 Table 11 -Location of Fog-Related Crashes ....... ............ ............ .... ..................... ...... .... ........... 20 Table 12-Fog Detection and Motorist Warning System Cornponents .......... . ............................. 34 Table 13 -Safe Driving Tips in Fog ................ ..... .. ............................ .... ....... .... ........... ........ ..... 40 LIST OF FIGURES Page Figure l-Monthly Distribution of Fog Related Crashes ............ ....... .......... .. ........... .. .......... ... ...... 8 Figure 2-Annual Fog Crash Rates ............. ..... ............... ... ............. ............... ................ ............... 12 Figure 3 Fog-Related Crashes in the Tampa Bay Area ...... ..... ........... .......... ............ ....... .. ....... .. 12 Figure 4 California Public Brochure for Fog .. ..... ...... ...... .................... ............ .. .. .................... 38


I. Executive Summary In February 1997, the Center for Urban Transportation Research (CUTR) was retained by the Florida Department of Transportation-District Vli Office to conduct a four-month investigation to determine: (I) the extent of unique and recurring patterns of fog and fog-related incidents in the Tampa Bay area (defined as Hillsborough and Pinellas counties); and (2) suitable counterme

campaign (to assure driver behavior is uniform in times of limited visibility), only Califomia has followed through in this endeavor. This report recommends and descn'bes a focused driver awareness campaign as the most cost-effective measure to reduce fog-related crashes, since the Tampa Bay area exhibits no particular fog-prone or fog-crash-prone areas. This awareness campaign should share information related to the fog seas on, fog crash history ;and driving tips in fog D. Backg round On December 27, 1996, at 11:30 a.m., a fug-related incident occurred on the Sunshine Skyway Bridge involving a 54-vehicle incident in both travel directions. This single event, although very uncharacteristic of historical fog-rela ted cmshes in the Tampa Bay area, piqued local interest and concern about fog detection and motorist warning systems tbat may be needed for the area (Hillsborough and Pinellas counties). Fog-related crashes, like crashes in general, are difficult to predict but may exhibit some tendencies associated with their occurrence. It has been generally concluded from National Transportation Safety Board (NTSB) investigations of major fog incidents that fog-related crashes result because drivers have not maintained uniform reduced speeds during times of limited visibility. However, just because drivers do not maintain uniform reduced speeds during period s of reduced visibility does not guarantee that a crash will occur. For example, according to FOOT, 5,700 total vehicles successfully crossed the Sunshine Skyway bridge in heavy fog conditions on the morning of the December 27th. Dense fog is a threat to the safe and efficient operation of motor vehicles Attempts are being made to prevent, abate, and disperse fug and to improve visibility and guidance through fog. Restricted driver visibility due to fog and its relationship to safe traffic operation, particularly on high-speed freeways, has been a national concern. However, it is important t o note that in Florida, from 1987-1995, the percentage of fog-related crashes to all crashes was 0.32 percent.' This statistic includes only crashes where fog was the primary environmental contributing cause. According to 1994 F ARS data, fog weather conditions existed in 1.6 percent of all fatal crashes nationwide. Compared to the1994 national average, Florida was 2.2 percent and South Dakota (having the highest percentage) was 5.0.2 Although fog crashes account for a 1 Florida Depanment of Highway Safety and Motor Vchloler, Office ot MlliUlgcmcot and Plaruting Services. Tlafllo Crashil1W>uo. 2 1994 Fltlal Accident Reporting System (FARs) D1tla, National Center for Statistics Analysis. 2


relatively small portion of all crashes, when fog was a contributing cause or the prevailing weather condition at the time of fatal crashes, they can involve many ':'ehicles in a chain-reaction pileup which attracts much public attention. These poor visibility conditions increase stress on drivers and reduce their ability to react appropriately to sudden changes in roadway and traffic conditions. Two very important aspects of fog crashes needed to be determined. First, the extent of unique and recurring patterns of fog and fog-related incidents in the Tampa Bay area were not fully documented. Second, suitable measures being utilized throughout the country to systematically and effectively detect fog and fog-related incidents and warn motorists in real time of these conditions were not known. Consequently, the Center for Urban Transportation Research was retained by the Florida Department of Transportation-District VII Office in February 1997 to conduct a four-month investigation to determine a basic definition of these aspects. ill. Purpose The purpose of this evaluation is to investigate and define the specific Tampa Bay area conditions for fog and fog-related crashes that may exist and recommend an area-wide plan based on these findings to ensure that drivers react more consistently and safely during times of limited visibility. This recommended plan will focus on the most appropriate techniques for detection, warning and related driver education and awareness programs. This report is structured to address four primary questions: (1) What are the recurring patterns (if any) of fog and fog-related crashes in Tampa Bay? (2) How does the rate of fog crashes in Tampa Bay compare with other Florida counties? (3) What are other states doing (in general) to address fog-related incidents? (4) Which countermeasure technique, or combination of techniques, would be justified for Tampa Bay given the .findings of(l) and (2) above? IV. Meteorologjcal Data Review Fog is one of the most serious meteorological limitations to visibility. The extreme variability of fog, especially in its density and location, make it difficult for motorists to perceive and react quickly. Fog can affect both day and night driving conditions because light, both natural and manmade, is retro-reflected, (refracted and deflected by the water droplets of the 3


fog) and will veil objects from sight. Fog is measw:ed by visibility in mile, and is considered severe (or "heavy") when visibility is 114 mile or less. If this condition persists for at least several how:s during the day, a heavy fog day is recorded. According to TouchWeather Wisdom, the foggiest location in the U.S. is located at Cape Disappointment, Washington, at the mouth of the Columbia River, with ail average of 106 heavy fog days per year. Eastport, Maine is the foggiest area on the eastern U.S. coast with 65 heavy fog days annually. Elkins, Virginia is the foggiest inland area in the U.S., with about 81 days annually with heavy fog. Many people assume that the San Francisco Bay area gets a lot of fog, but it averages only 18 heavy fog days a year (slightly less than the average for the Tampa Bay area at 22 heavy fog days a year). Informal surveys conducted with the National Weather Service in Ruskin, Florida, and several local meteorologists provided the basic characteristics offog and fog forms. Fog can be defmed as a cloud in contact with the ground composed of tiny droplets of water or ice crystals. These droplets form spherical shapes, and their diameters may range fro m two to I 00 microns Fog usually forms in two ways: (I) by air cooling to its saturation point, and (2) by air parcels mixing with different temperatures and humidities. There are fow: prevalent types offog. The earth's radiational cooling produces radiation jog. It forms when drier air has overlain a layer of moist air near the ground during late fall and winter The moist l ower layer, chilled rapidly by the cold ground, quickly becomes saturated, and fog forms. When a high pressw:e system becomes stagnant over an area radiation fog may form on many consecutive days. This fog is also known as valley fog" because the cold, heavy air drains downhill and collects in low-lying areas. Advectwn fog is the fog that arises from the movement of humid air over a surface that is already cool. This type of fog is most prevalent in the regions of Pacific coasts, and the southern and central United States and tends to form over large grassy areas. Upslope fog forms when moist air flows up along an elevated plain, hill or mountain. Evaporation jog or Sea fog forms when cold air move s over warm water. When rain drops fall through into a cold layer, these rain drops will be under a high vapor pressw:e and the water from rain drops evaporate. When the cold air becomes sufficiently moist, fog forms Sea fog is much thicker than advection fog and takes longer to dissipate when it comes off the warm gul f waters. Radiation, advection, and evaporation (sea) fog are all common to the Tampa Bay area. Fog is something we have to learn to cope with. Basically if we did not have cooler air masses (or cold fronts) moving over warmer land and water, fog would not form The U.S. Air Force has experimented with fog dissipation on a small scale with silver iodine generators (which 4


"rain-out" the aix's moisture), but success of this project has not been documented. Also, large fans have been used to stir-up the air over small areas, but not on a larger scale. Consequently, when fog does form, real-time information on the presence and density of fog is necessary for effective traffic control. Presently, fog-related information is available from several sources. For exatnple, the National Oceanic and Atmospheric Administration (NOAA) Weather Wire Service, using a national satellite-based information gathering system, collects and reports all types of . weather data. Also, NOAA's Radio Network System offers routine weather information, incl uding dense fog advisories that reaches about 90 percent of the U.S. population.3 It is interesting to note that for two days prior to the December 27, 1997 Skyway pileup, the U .S. Weather Service office in Ruskin, Florida had been predicted dense fog for Hillsborough and Pinellas counties. In the Tatnpa Bay area, hourly weather updates are provided on the Internet via the Florida Weather Center @ http://www. weathercenter.com, a free informational service provided by WFLA-TV News Channel 8 meteorologists. The National Climatic Data Center in Asheville, North Carolina collects detailed historical local climatological data from one collection point (Tatnpa International Airport) in the Tatnpa Bay area based on hourly averages for all days in the month and three-hour observation intervals for each day in the month. Fog prediction can be very difficult because of the variability in density, location, development and dissipation rates, and area of coverage. According to the National Weather Service forecasters in Ruskin, "there is no particular favorite location for fog to form in the Tatnpa Bay area." Further, the National Transportation Safety Board (NTSB) has concluded that "although weather forecasts may alert authorities to the possibility of fog formation, they are not sufficiently accurate, comprehensive, or timely to predict that fog will form in a specific area."' Though meteorologists often can accurately forecast the initiation of conditions necessary for the formation offog, the expected fog does not always appear, or it may appear under conditions that are not ideal for fog formation. WFLA-TV News Chief Meteorologist David Grant concurs with the National Weather Service and NTSB in their conclusions, but as previously stated, the ideal conditionS for the formation of fog can be identified. Mr. Grant offers the following four-part "formula" for the most favorable conditions leading to the fonnation of fog: 3 'Reduced Visibility due to fog on ffi8hway," NCl!RP Synthesis 228. 4 "Special Public Hearing on Fog Aec.idents .on Highways: N&lional Transportation Safety Board. 5


(1) Air temperatures between 40-60 degrees F (2) Sufficient moisture content (dew point close io air iemperature, high relative humidity) (3 )Calm to fairly light winds (less than 2 mph) (4) Clear skies (since ground will radiate more readily) These conditions are generally known to simultaneously occur primarily during the months of December, January, and February. During these months, the Tampa Bay area generally has cool nights with. little or no.winds. This.typical "fog season" for the Tampa Bay area can also be characterized by. examining summary data from the National Climatic Data Center and Florida Department of Highway Safety and Motor Vehicles (DHSMV). Table I below summarizes average readings for climatological data (midnight 7am) for eight selected days during the typical "fog season" when fog was recorded. The average of these values generally coincides with the previously mentioned ideal conditions for fog formation. Additionally, Figure I illustrates the monthly distribution of the fog-related crashes recorded by the Florida DHSMV for the period 1987-1995. Almost 60 percent (57.77 percent) of all reported fog-related crashes occurred during the months of December, January, and February. During the months of December and January alone, 43 percent of the crashes occurred. 6


Table 1 Selected Climatological Data During "Fog Seaisl!n" Visibility for the o bservatio n period prior to fog being recorded. Source: National Climatic Data Cent e r Tampa International Airport, Asheville, Nortb Carolina. 7


.... noo "" .... .... I ... "' I ... ... ... ... ,.. ... Sol,ll'oe: Figarel Monthly Distribution ofFoc Related Cruhes (1987) "'.......;::_ I -$Dlho-D

V. Crub Data Rmew This section contains an overview of all motor vehicle crashes from 1987-1996 in Hillsborough and Pinellas counties in which fog was a primary contributing environmental cause of the crash. Data for the analysis were obtained from long-fonn crash reports contained in the Florida Traffic Crash Database and were provided by the Department of Highway Safety and Motor Vehicle (DHSMV), Office of Management and Planning Services. Crashes are recorded on a long-form and entered into the database only when they involve death, personal injury, drivin g while under influe nce of alcohoVor chemical/controlled substances, bit-and-run, or significant damage to the vehic l e that requires removal from the crash scene. The inoperable vehicle requirement was dropped several years ago from long-fonn crash reports. For analysis piUpOseS, 10 years of crash record data for Hillsborough and Pinellas counties were combined and frequency distributions were computed to identify crash, driver, vehicle, and roadway level characteristics. However, 1996 crash record data were incomplete due to the lag time between when the crash report originates at the local l evel and when that report is entered into the statewide Traffic Crash Database. As such, 1996 data were excluded from summary discussions The following sections highlight the results of the data analysis. Crash Data Analysis This section contains information on the incidence of fog-rel ated motor vehicle crashes, the number of fatalities and injuries actions committed by drivers that contributed to the crash, the time of day that fog-related crashes are likely to occur, and the severity of crash and injury in fog-related crashes Fog-Related Crashes, Injuries and Fatalities: Hillsbcrough and Pinellas Counties and Florida. 1987-1995 Statewide, 6,323 fog -related motor vehicle crashes occurred from 1987-1995 Of these crashes, 829 (13 percent) occurred in Hillsborough and Pinellas counties. Fog-related crashes peaked in 1989 when, statewide a total of 1,1 S I crashes were reported. That number dipped to 462 in 1991, the lowest number of fog-related crashes recorded during the nine-year period. Over that same period 300 people were killed and 7,169 were injured on Florida's roadways in motor vehicle crashes in which red uced visibility (fog) was a contributing factor to the crash. Twenty-nine fatalities occurred in Hillsborough and Pinellas counties and another 812 people were injured. The overall percentage of fog-related crashes to all motor vehicle crashes in 9


Hillsborough and Pinellas counties is approximately equal to the statewide percentage over the nine-year period (0.30 percent compared to 0.32 percent). A summary of 1987-1995 fog-related crashes, injuries, and fatalities both statewide and in Hillsborough and Pinellas eounties is shown in Table2. Table2 Motor Vehicle Crashes, Fog-Related Crashes, and Crash Severity, Hillsborough and Pinellas Counties and Statewide, 1987-1995 Source: of Highway Safety and Motor Vehicle$, Oflioe of Management and Planning S..Vices, Traffic Crash Database. Two other key aspects of crash level data also have been summarized for purposes of this report: geographic location and rate. Hard copies of all fog-related crashes (1987-1996) were reviewed to depict each crash report site on a geographic base map. Only those crash reports with legible locations have been incoiporated into Figure 2, Fog Related Crashes in the Tampa Bay Area. A total of 809 crash report sites were plotted on the map. Note that fog-related crashes over the last I 0 years have occurred throughout the entire area and that there is no particular fog-crash-prone area. The scope of this evaluation did not include a comparison of the spatial distribution of fog-related crashes to all crashes. What appears to be clustering of crash sites (e.g., Plant City area, north along U.S 41, Gandy Bridge) are crashes at different locations spread out over multiple years. These general areas can be utilized for future fog detection/ warning system evaluation. 10


Annual crash rates have been calculated per 10 million daily vehicle-miles traveled (VMT). Annual VMTs for all public roads, by county, were provided by FOOT's Transportation Statistics Office. Figure 2 illustrates the trend In this rate for the period 1987-1995 for Pinellas and Hillsborough counties compared to the statewide average. Hillsborough County has annually ranked above the statewide average, with its highest ranking reached in 1992 at 16th among Florida's 67 counties. The abrupt drop in the 1991 crash rate for Hillsborough County was due to a 71 percent drop in fog-related crashes with only a 10 percent drop in vehicle-miles traveled. On the other hand, Pinellas County has annually ranked below the statewide average, with its highest ranking also reached in 1992 at 47th among Florida's 67 counties. Fog crash rate calculation sheets for all Florida counties, by year, are contained in the appendix of this report. One additional major finding can also be reached by review of these crash rate tables in the Appendix; with the exception of only two years, Hillsborough County has reported the greatest number of fog-related crashes in the state. 11


. Figure'l ADDUII Fog Crash Rates ,. M .. ----J __ .. A .. .. .. "" .... .... .... " "" Noto: Cn.sh Rate indicates tho ann11al number offog-relate:bose 19S1199S. 12


Priver Contributing Causes in Fog-Related Crashes Although fog is the prinWy environmental contributing cause in these crashes, drivers often commit errors that also contribute to the crash. However, in 42 percent of the fog-related crashes, drivers were not issued citations for improper driving techniques (see Table 3). lbis is, in part, because law enforcement officers or other motorists must witness the infr!'ction, which is extremely difficult under reduced visibility conditions. In crashes where drivers received citations, 19 percent contributed to the crash by driving carelessly, 9 pereent failed to yield the right-of-way, and 5 percent exceeded safe speed. Table 3 Most Common Driver/Pedestrian Contributing Cause in Fog-Related Crashes, Hillsborough and Pinellas Counties, Florida, 1987-1995 Cause .'\umber of Times Cited Percent Cause Cited No improper driving/action 592 41.7 Careless driving 267 1S.S Failed to yield right-of-way 125 s.s Exceeded safe speed limit 71 5.0 Q1het" 364 25.7 Total 1,419 100.0 *Drivers can be cited for more than one contributing cause. Otbu includes; improper backing. improper Jane c.bange, imprope.t tum, alcohol-under influence, aloohol &. drugsunder influence, followed too closely, disregarded Uaf!ic signal, disregarded stop sign, failed to maintain pass in& drove left of center, exceeded stated speed obstructing disregarded other traffic c:ontro1. driving wrong side/way and all othc. r. Source: Departm.ent of Highway Safecy and Motor Office ofManagement and Planning Services, Traffic Crash Database. Time Table 4 shows all fog-related crashes that occUited in Hillsborough and Pinellas counties during 1987-1995 by the time of day when the crash occUited. Almost one half (48 percent) of the fog-related crashes occurred between 6 a.m. and 9 a.m., with the highest concentration of crashes occurring between 6 a.m. and 7 a.m. More than one-third (37 percent) of the fog-related crashes happened between midnight and 6 a.m. However, crashes during this time period tend to be more evenly distributed. Because the majority of fog-related crashes occur during the a.m. peak commute period, public service announcements (PSAs) promoting safe driving techniques in fog could be aired during this time. 13


Table4 Fog-Related Crashes by Time of Day; Hillsborough and Pinellas Counties, Florida, 1987 1995 Time of Day Number offog-rclatcd c rashes Percent of fog-related crashes Midnight to 2:59 am 146 17.6 3 am to 5:59am 161 19.4 6 am to 8:59 am 399 48.1 9 am to 3:59pm 25 3.0 4 pm to 7:59pm 13 1.7 8 pm to 11:59 pm 61 7.4 Unknown 24 2 .9 Total 829 100.0 Source: Department oflllghway Safety and Motor Vehicles, 01'11oe ofManagemC31t and Planning Svices, Traffic Crash Database. Crash Injury Severity in Fog-Related Crashes Crash injury severity indicates the overall injury of the crash and is defined by the most severe injury to any person involved in the crash as perceived by the investigating officer. The crash injury severity for fog-related crashes in Hillsborough and Pinellas counties ranged from 37 percent of the crashes resulting in no injuries to 40 percent resulting in some type of non incapacitating' or incapacitating injurl. Possible injuries7 were noted in 20 percent of the fog related crashes. A total of24 (3 perc ent) of the crashes resulted in fatalities. Thus, some crashes resulted in more than one fatality. A summary of the crash injury severity in fog-related crashes is contained in Table 5 5 Non-incapacitating iojwy is defined as any visible injuries such as bruises, abrasions, Umping. etc. 6 Jncapaeitating irljwy is defined as any visible signs of injwy from the crash and person(s) had to be carried from the crash scene. 7 PO$Sible injwy means no vis ible signs of injury, but complaint of paiD cr mc>mentaiy unconsciousness. 14


TableS Crash Injury Severity in F o g-Related C rashes; HUisborough and Pinellas Coui1tles, F i orida, 19871 995 In jury S('HI i\lmt St.-.ete lnJUfJ Perccnl of I or.d No iqjury 304 36.7 Possible injmy 166 20.0 Non-in<:apaciwing 228 27.5 Incapacitating 107 12.9 Fatal 24 2.9 Total 829 100.0 Soon:e: Department otHI;J\way Sattcy and Motor Vcbkk s. ofManagemcnt and PIIMlna Services, T.raffi4; Crash duabue. Driver Data Analvsis This section contains infonnation on the age and place o f residence of drivers involved in fogrelated motor vebi c l e crashe s Age Table 6 shows the distribution of driver age group s involved in fog-re lated crashes in Hillsborough and Pine llas counties. Overall, younger and midd le-aged drive rs tend to be involve d in fog-relat ed crashes more often than older drivers A total of 411 (33 percent) of the drivers involved in fogrelated crashes were age 20 to 29 years, while 23 percent of the drivers were age 30 to 39 years. In part, over-representa tion amon g young and middle-aged drivers ma y be a function of the time of da y fog-related acciden ts typically occur (e. g., peak a.m. school and work cornmut e periods) and the gen eral lack of driving exp erienc e in low visibility con ditions. Thus, these resul ts suggest that traffi c safety education c urriculum in high sc hools and universities might include instruction on safe driving techniques in reduced visibility conditions 15


Table6 Age Distrlbu tioo ofDrivers lav o lved in F og-Re l a ted Crashes; BDis. horough a nd Pinellas Coan tles, Flo rida, 1.987-1995 nn,tr t: of in .\ge Croup '%, l>nH;r in Age <.ruup 14 to 19 years 171 13.9 20 to 29 years 411 33.3 30-39 years 218 22.$ 40-49years 181 14.7 $0-59years 95 7.7 60+years 84 6.8 Unknown 14 1.1 Total 1,234 1 00 .0 So= Deportment ofR!gbway Safely and Mooor Vehicles, Office and P1anoillg SetviatWs<. Residence Status A total of 1,019 (83 percent) of the drivers iiiVolved in fog-related motor vehicle crashes were residents of the county in wbicb the crash occwred (see Table 7.) These results indicate that most drivers were familiar with local roadways and conditions. It may be that drivers wbo are more familiar with roadway conditions drive less cautiously in adverse wea th e r conditions. However, any co rrelation betw een roadway familiarity and driving habits under reduced visibility condi tions cannot be assessed. These results do show that few drivers involved in fog -related crashes in Hillsborough and Pinellas counties are residents of other states or countries. Place of Rt,idrnccCounty of c:msh Elsewhere in state Non-resident of state Foreign U nknown NIA Total Table7 Driver Residence Status In Fog-Related C r ashes; Bill s bor ough and Pinella s Counties Florida 1987-1995 :"jumher of dri\'l'fS Percent of tlrh crs 1,019 82.6 165 13.4 36 2.9 6 0.5 3 0.2 s 0.4 1,234 100.0 Souree: Departm<>U ofHii)lway Safely and Motor Vehicles, Office of Management and Planni ng Services, Traffic Crash Database. 16


Vehicle Data Analysis This section contains information on the type of vehicle involved in fog-related motor vehicle crashes and the movement of the vehicle when the crash occurred Vehicle Type Table & contains a breakdown of the type of vehicles involved in fog-related crashes in Hillsborough and Pinellas counties over the past nine As expected, the largest percenlll.ge of vehicles (73 percent ) in fog-related crashes were automobiles and passenger vans. These results reflect the higher percentage of registered automobile and passenger vans relative to other vehicle types. Approximately one quarter of all vehicles (22 percent) involved in fog-related crashes were trucks. Of the 267 trucks involved, 70 percent were pickup trucks and 30 percent were medium and large trucks. TableS Type ofVebide* Involved iu Fog-Re lated Cnsbes; Blllsboro11cb aad Pinellas Couatia, F l o .rlda, 1987 \ d1iclt. pe :\umlH:r o fYehiclr' Penent of\ chidl" Automobile and passenger 900 72.9 Light truck (pickup) 1&7 15.2 Other truck* so 6.5 Motorcycle 21 1.7 Bicycle 14 1.1 Law enforcement vehicle II 0.9 Bus 8 0.6 Motor home (RV) 5 0.4 Taxicab 4 0.3 Other 4 0.3 Total 1,234 100.0 Modificallonsto vehicle cataa&ma Savi

Vehicle Mqvement The majority of vehicles involved in fog-related motor crashes were traveling straight ahead when crashes occur (see Tahle 9). Of the 1,234 vehicles involved in fog-related crashes, 68 pereeot of the vehicles were traveling straight ahead; 16 percent of the vehicles were slowing, stopping, or stalled, and I 0 percent of the vehicles were turning left at the time of the crash. Table 9 Movement of Vehicles Involved in Fog-Related Crashes; Billsboroug)land Pinellas Counties, Florida, 1987-1995 \ chicle J\lovcmtnt Number uf \ l'11idcs Percent of \'chicles Sttai ght Ahead 834 67.6 Slowing/Stopped/Stalled 197 16.0 Mmdng Left TUm 121 9.8 AD Other 82 6.6 Total 1,234 100.0 'Other includes backing, making rigbttum, c.banging lanes, ent<:rins/leaving parklna space, properly parked, Improperly po.rlallmeOt offfigbway Saftcy and Motor Vebieles, Office ofManagemeot and Planning Servi<>es, Tnllie era.b Dmbase Roadway Data Analysis This section contains infonnation on the type of roadway system where fog-related motor vehicle crashes occur as well as jurisdictional location where crashes tend to occur. Type of Roadway System When Fog-&lated Cra.shes Occur Table I 0 contains a summary of the type of roadway systems where fog-related crashes are more likely to occur in Hillsborough and Pinellas counties. Most of these crashes tend to occur on county and local roads, the most frequently traveled roads within the counties A total of 32 percent of the fogrelated crashes occurred on local roads; 30 percent occurred on county roads. A total of 189 of the fog related crashes (23 percent) took place on state roads. 18


R oadway Type Local County Swo )1)_ .. u.s. Other Total Table 10 Roadwa y Type Where Fog-Re l ated Crashes Occurred; ffillsborough and Pinellas Co1111tles, Florida, 1987-1995 Number of Crashe s Percent of Crashrs 267 32 2 246 29.7 189 22.8 64 7.7 41 4.9 22 2.7 829 100.0 S<>urce: DepBrtment ofHiibway Safety and Motor Vehicles. Offiee ofManqcment and Planning Services, Traffic Crash Datnbase. Location of Fog-Related Croshu Historically, crasb locat ion bas been coded on the crash report as either rural or iuban 8 In 1993, an additional field was added to the crash report that defmed tbe area as business, residential, or open country to more accurate ly reflect the environmental location of the crash. As Table I I indicat es tbat the majority of fog-related crashe s (66 percent) in Hillsborough and Pinellas counties occurred in rural l ocations. An examination of the data from 1993--1995 shows that 43 percent of fog-related crashes occurred in areas considered to be primarily business locations and 32 percent occurred in residen tial locations. One-forth of the fog-related crashes during this pe:ciod occurred in locations considered to be open country. Rural jndic:a.te$ tm.t the crash oeewred outside the dty limits or whhln the limits of a cUy with a populadon less than Z,SOO population. Urbon indicates th&llhc crosh occurred within lhe Umils of cities and certain otht police jurisdictions with population-s greater chan or equal to 2,SOO population. 19


Table 11 Locatioa of Fog-Related Crasbes; Hillsboroueb aad Pinelias Couatie s, tltorlaa, 1987-1995 I .uration of Crash Number of Crashes l'crccnt of Ctashcs -. ' ----' .. . --... ------_. ________ ""'' ------..... -. -----, ', ,' ... J \ ',' .... ,. __ . ,-.: '". ';: -__ .....__._ _____ --. . __ ....__ ______ ... . Rural 549 66.2 Urban 280 33.8 Total 829 100.0 Primarily business 74 43.0 Primarily residential 55 32.0 OpCII country 43 25.0 Total 172 100.0 Source: Dtpa.rtment of Highway Safety and Motor Vehicles, Office of Management and Planning S

VI. Technologies Visibility Detection Tampa International Airport is !lie principal reporting station in the Tampa Bay for the National Climatic Data Center The other general aviation aiiports and t e levision stations in the area have minimal weather and visibility equipment. The Port of Tampa has one visibility sensor about five miles west of Egmont channel and seven meteorological sensors (temperature, .atmosp h eric pressure, wind speed, and wind direction) position e d throughbut Tampa Bay. There is also a rain detector on Clearwater Beach and wind monitoring equipment on the Sunshine Skyway Bridge. Therefore, the extent of visibility monitoring is conducted from onl y several point sources in the area, nothing within major travel corridor rights-of-way. Real-time information on the presence and density of fog is important for carrying out countermeasures because any time gap between the onset of fog and the initiation of safety measures could be critical. Such information can be obtained by deploying fog and weather detection devices. Fog sensing devices have been in use at aiiports, waterways, and on some highways. There are three types of instruments available to measure visual range on a continual basi s These devices are readily available and have a wide price range. They are categorized as transmissometers, back scatter sensors and forward scatter sensors. Both forward and back scatter sensors can forecast the visibility conditions over a small volume of air, becoming "point detectors." In a transmissometer, a projector transmits a known amount of light toward a detector usually set at a distance of about 1,000 feet away. Primarily used at aiiports, these instruments are costly, heavy, and require a long and accurate aligiunent. These instruments are not suitable for highway applications because of the problems involved in their installation. For example, source and receiver of a light source have to be placed in a clear line-of-sight (minimum of I ,000 feet apart) which cannot always be m et on highways, and these devices are also very expensive, ranging from $10,000 to $15,000 each The optics used in transmissometers also require frequent maintenance due to normal highway air quality environment. In a back scatter senso r the light source and receiver is pointed in the same direction and positioned in such a manner that light scattered back can be measured. A large amount of light scattered back indicates dense fog. Back scatter devices are one of the oldest technologies in this field and cannot differentiate among various poor visibility conditions like fog, snow, or rain 21


drops. Another disadvantage of Ibis device is the variation in the amount and direction o f back scattered light. The forward scatte r visibility sensor is an active e l e ctro-optical instrument that detennines visibility by measuring the optical extinctio n coeffi cient of a beam of light as i t passes through a known v olume of ai r Particles in air such as fog, rain, or snow affect the extinction coefficient. This value is th e n transmitted to an external computer in its unaltered form or translated into an equivalent visibility in miles or kilometers. The sensor projects a beam of light into a receiver that measures fog and light scattered forward into a receiver is measured. Although new, Ibis sensor is competitive in accuracy reliability, and cost. Its lightweight, compact, easily mountabl e structure make it ideal for highway applications. The cost of these sensors range from $5,000 to $8,000 The compact s ize and simple al ignment require ments make the forward and back scatter sensors practical for highway applications. In these sensors, the source and the receivers of infrared light are placed at distances less than one meter apart thereby avoiding the line-of-sight problems. However, there are no established standards or precedents on the number of sensors required and ideal spacing configurations. This is primarily due to the limited information and evidence available on the formation of fog and its variability . It is known that fog is generally not "site spe cific" and varies from place to plac e Thus, it is difficult to suggest specific gui d elines on number and spacing req uirements. The information on fog can also be obtained by installing weather stations in fog-prone areas. Meteorology of fog shows that fog formation will be accompanied by some weather parameters like wind speed, temperature, humidity, and dew point. Weather stations equipped with day/night detectors, wind speed sensors, temperature/relative humidity sensors rain gauges, and barometric press ure sensors provide information to monitor and forecast fog formation. These weathe r stations are also useful to correlate weather parameters with the historical valu es, and, hence, it may b e possi bl e to arrive at id eal conf'l!urations for fog detecto rs. Clos ed circuit television (CCIV) cameras are also being utilized as a viable mechanism for monitoring and confirmin g adver se weather conditions. Various facilities in United States have deployed or are deploying different types of fog detection devices, but, many areas are still relying on manual observatio n of fog. The Caltrans "Highway Fog: Visibility measures and Guidance Systems, William H. Heiss, NCIIRP Report 171. 22


Meteorological System in the fog-prone Central San Joaquin Valley of California is equipped with high perfonnance sensors and data a Cquisition equipment installed at nine separa t e locations. The device provides real-time weather and visual range data for a larg e monitoring area. They include remote sensor assemblies consisting of pavement sensors, forward scatter fog sensors, wind speed and direction detectors, barometric pressure recorders, rain gauges etc., and a central processing unit A master computer uses the data to assess conditio ns and provide reports of special weather conditions to drivers within the monitored area.10 The cost of the entire project was more than S3.6 million ($1.32 million for California Department of Transportation CAL TRANS and $2.35 million for California Highway Patrol CHP).11 Louisiana is relying on duty personnel to observe and monitor the highway facilities during fog and pass the information to control towers. However a recent accident on a five-mile bridge on I-10 forced the LDOTD authorities to study the feasibility of fog detection and motorist warning technologies. Their study recommended not to install any detection technologies like fog sensors and cameras, estimated to cost about $330,000 and $500,000, respectively. Their recommendation was based on the maintenance, communication, and standardization problems they perceived. Their decision was also based on an FHW A evaluation study on sensor technologies indicating the discrepanci es in their accuracy ranges. The LDOTD study alsO concluded that the best and most effective system would be to rely on law enforcement for fog detection. 12 South Carolina installed weather monitoring equipment consisting of fog detectors and weather stations. The system resulted from a federal court action requiring the South Carolina Department of Transportation (SCD01) to provide a plan for mitigating the effects of fog. The court action was a result of concern about the effects fog cn:ated by a paper mill near the Cooper River bridge in Charleston. (It could not be determined whether the paper mill was held liable for any mitigation costs.) The system is equipped with five forward scatter type fog detectors at 500-foot interval s. The system also has a weather station to detect wind direction, wind speed, temperature, and humidity These devices provide information to a data r ecorder and a central computer to correlate the prevailing field conditions with a set of preselected parameters to 10 User Manual on "Cal1rons Mcuon>lo&ical System," Rt:port I 11, Qualimcuico,IJ>:, I 997. 11 "Stral

determine the appropriate counteaneasures of reduced visibility. During 1960s, the New Jersey 1\unpike Authority (NIT A) contracted with a private weather forecasting service to provide three daily forecasts and additional forecasts when foggy conditions are expected. For a short period oftime during the mid 1970s, the turnpike opted for a laser system for fog dete ction. In the middle of the 1970s the turnpike opted for a laser system. However, installation problems, coupled with components failure and difficulty in finding replacement parts forced the turnpike to abandon the project. Instead, NITA sought off-the shelf detectors proven by other agencies and purchased two fog detectors and complete weather stations." In 1993, another fog detection and motorist warning system was install ed on a river bridge on 1-287 in New Jersey." The system developed for the 2,000 -foot bridge was equipped with a forward scatter fog sensor known as Fog Sentiner>' FSA Series visibility sensor, designed specifically for highways Built-in circuitry can activate warning instrumen ts like signs and lighting systems. In the present case, it was designed t o activate a light guidance system. The cost of the forward scatter senso r was about $5,600. H owever, the principal form of fog detection continues to be the personal observation by the State police. The Idaho Transpo rtation Department is coqtinuing the development and testing of three types of sensors for measuring visibility and weather: Scanners, HANDAR, and LIDAR, provided by three individual companies. Scanner is provided by Surface Systems Inc. The HANDAR system is provided by HANDAR Corporation and it includes one portable remote environmental monitoring system that measures weather condition, and one visibility sensor. Both Scanner and HANDAR are typical forward scatter detectors, and LIDAR is a laser employed visibility detector provided by Santa Fe Technologies. The detector has a single visibility sensor and is incorporated with advanced laser technology recently deve loped at Los Alamos National Laboratories. The primary differenc e between LIDAR and Scanner or HANDAR is that the LIDAR system is capable of measuring visibility conditio ns over a large area. These sensors are used not only detecting fog but also other poor visibility conditions like snow, blowing dust etc which are predominant in Idaho. HANDAR is considered the most cost-effective, and LIDAR uses the latest laser technology The costs are expected to be around 13 NCHRP Syotbcsis 228. Telephone oo.....ation with Fr!tlk Dellarossa, FHW A Divisional Office, New ]etSey. 24


$15,800 for HANDAR and $75,000 and LIDAR." Following three severe chain reaction crashes (in 1978, 1979, and 1990) on I-75, Tennessee has developed a fog detection system. The I-75 system covers a 19-mile section of the highway identified as the fog-prone area. The system continually monitors the climatological and visibility conditions along the three-mile highway section with a history of severe fogging events. Eight forward scatter fog detectors integrated with two weather stations monitor visibility across the fog area. The weather stations measure temperature, wind speed, wind direction, and dew points. The information is processed by using the Management Information System for Traffic (MIST 2.0) developed by Farradyne Systems, Inc. Climatological threshold criteria are being used to alert the operators in the central control center that a response is warranted.16 The system was set up to operate in four different pre-programmed visibility scenarios for operating variable message signs: (1) clear--no visibility deterrent; (2) moderate--moderate visual impairment; (3) severe-severe visual impairment; and (4) critical--critical visual impairment. Depending upon the visibility scenario, various messages have been pre-programmed for displaying on variable message signs. The entire project cost about $4 5 million. The Alabama Department of Transportation also is planning to install a fog detection system on a seven-mile flat sea bridge on I-10 near Mobile. This system will be equipped with seven forward scatter fog sensing devices and one weather station with several weather instruments that can detect wind speed, wind direction, temperature etc., These weather and fog detection devices will be integrated with other motorist wamiilg technologies The Georgia Department of Transportation and the Georgia Tech Research Institute are Telephone conversation with Fred Kitchc:ncr, Projetl Manager, CH2M Hill, reprdlng Idaho study 16 Telephone conversation with Dave Cox;FHWA Divisional Office, Tennessee, Florida. 17 Telephone c:onversation with Paul Watson, ALDOT Electrical Engineer. 25


developing a fully-automated fog detection systel_ll along the heavily-traveled section of 1-75 north of the Georgia/Florida border. The $3 million system is equipped with 19 forward scatter type fog sensors and several other types of weather monitoring devices including precipitation, wind, hwnidity, and temperature measuring instrwnents to monitor the visibility conditions over a 2-rnile section of the 13-mile long fog-prone section of the highway. The primary objective of these. weather instruments is to detect the poor visibility conditions caused by conditions other than fog, such as smoke from agricultural burnings. These conditions are also used for study various weather parameters that contribute to fog .formation. Information from these devices is sent via buried telephone lines and the system is also designed for transmission through fiber optics in the future." The fog sensors are expected to cost about $5,000 and the integrated weather stations under $6,000. The problems of poor visibility conditions posed by fog are not limited to United States alone, and several European countries are also making efforts to counter the adverse impacts of foggy conditions. Project DRNE in the Netherlands has proposed to install an integrated system of nephelometers to assess road visibility. The nephelometers measure the physical structure of the clouds, including their concentration, and the shape of cloud particles. PROMETHEUS' research program in Europe has developed a visibility monitoring system based on infrared laser beams (similar to the detector being tested in Idaho). The back sca tter signals from the beam are processed to derive the visibility range. Motorway 25, which circles the city of London, is equipped with fog detection technologies to detect and forecast poor visibility conditions. The Automatic Fog Warning System (AFWS), equipped with backward scatter sensors, is designed to help drivers by providing real-time information on weather conditions. Incident Detection and Motorist Warning The National Transportation Safety Board believes that "the ITS program offers a unique opportunity to develop and carry out limited visibility traffic control measures. Traffic flow detectors, automatic message and vehicle speed control systems, and radar vehicle detectors to warn of preceding objects, such as other vehicles, are all appropriate candidates for ITS projects/'19 18 Telephone eonvusatlon with Dr. GillY Gimmestead, Georgia Tech Research Institute. 19 NCHRP Synthesis 228. 26


Reports describing various accidents that have occurred due to poor visibility conditions in United States show that non-uniform driving speed is the most cause of these accidents.20 They also show that drivers are observ ed to maintain different speeds and headways according to their individual perceptions about the conditions and risks, lackll,lg any specific behavioral guidance or warning systems. Previous experiments also proved the fact that driver's reaction tirne improves significantly with the provision: of warning signs.21 .These warning systems could be either pass i ve traffic control systems like fixed signs, raised reflectorized pavement markers, upgraded striping standards or active trafiic control systems with variable message signs, surveillance systems, speed loops, closed circuit cameras. Currently within the Tampa Bay area, there are no incident detection systems. Changeable message signs are installed at three locations in each direction approaching the Sunshin e Skyway Bridge. It is anticipated that a $2 niillion variable message sign system will soon be installed along I-275 approaches to Tropicana Field. Three surveillance cameras exist along State Road 60 east ofl-75 and 13 cameras exist along the Sunshine Skyway Bridge. The City of Clearwater has one portable camera that is transported from site to site as needed. Plans exist in Hillsborough County for installation of surveillance cameras at nine north Tampa intersections in July 1997. Although not extensive at this tirne, the foundation for an area wide surveillance and motorist warning system is beginning. traffic control features lik e fixed signs are useful for less adverse conditions and also serve as a backup for active control features. Generally, fixed message signs are used to identify fog-prone areas. However, these signs may not be very effective, because the traveling public may consider them to be irrelevant since they represent the prevailing conditions only for a portion of the year. Another disadvantage of fixed signs is that they also may have to be ffipped open manually during limes of poor visibility. An active motorist warning system is an integrated system of various technologies to perform different tasks. All these technologies can be operated, guided and controlled from a centralized traffic management center. Such technolo gies may include variable message signs (VMS), highway advisory radios, street lighting controllers, surveillance systems with CCTVs, lighted pavement markers (LPM), visual readout radars, barrier rail reflectors, and traffic flow measuring equipment. 20 'Hiibway Accident Report on 1-40 er.sbes, Arlcansas. 21 Highway Accident Report on 1-40 Crashes, Ark>nsas. 27


These technologies can be integrated with vis i bility detection equipmen t and other systems like weather monitoring centers, integrated nephelometer, and lmowledge-based expert systems, lpld can also be activated automatically from central traffic monitoring centers. It is also possible to classify the prevailing conditions into several classes, depending on the visibility conditions like potential fog, light fog, moderate fog, severe fog, critical fog, and, based upon the prevailing conditions, appropriate information can be flashed on VMSs. VMSs can also be used to inform drivers to tune to radios and other information sources to have an update on weather conditions, visl"bility standards, and road conditions. Detailed information on road and prevailing visibility conditions can be provided through portable highway advisory radio (HAR) stations The l ow-power A.M. band radios can be equipped with changeable and pre-recorded messages to describe the visibility conditions and guidance measures. Experiments have shown that variable message signs placed before HAR station alert motorists to tune to HAR. The HAR equipped with cellular capabilities (as being done in Tennessee) can be connected to a central management center so that appropriate messages can be transmitted according to the situation. Real-time detection of traffic flow characteristics is important, not only for decreasing the delay on the freeway and city streets, but also in prev enting secondary accidents .22 It can be achieved by deploying flow interruption monitoring equipment like inductive loops, radar detectors, beacons, ccrv surveillance systems, video imaging, and magnetometer, etc. Inductive loops are the most commonly used vehicle detector. However, the application of this detection method is not recommended for the facilities like bridges since it may cause some adverse effects on the bridge due to the installation of loops in the bridge deck. 21 Radar detectors are another type of device that can be used to measure traffic flow and speed. However, they need to be mounted over the lanes to get accurate information and this will require an extensive number of overhead structures. Video imaging is new technology developed for traffic detection. In this technology, computers are used to process the images produced by closed circuit cameras This method can be used to monitor both vehicular flow and 22 'Higbway Accid"'t Report on J-40 Cmbcs, Arbnsu. 21 'Fog Dotl:

speed. However, these technologies are susceptible to failures during poor visibility conditions. Magnetometers are very useful for monitoring traffic flow on bridges by mounting them within and beneath the bridge decks Another device in the research and development stages that is useful to counter the problem of non-uniform driving speeds is visual readout radar. By using this system, the speeds being maintained by motorists can be flashed on the visual radar unit followed by a VMS showing the prevailing visibility conditions It will also have a monitor to measure the speeds to figure out the effect of variable message sjgns .24 Surveillance systems like CCTV cameras can be installed on the roadways in the fog prone areas to verify operation of the signs, weather conditions, and traffic incidents. Each site is equipped with cameras with zoom, pan, and tilt capabilities, along with encoding devices to convert an analog camera output into a digital signal for transmission over telephone lines. These systems are capable of providing the visual information necessary to select appropriate VMS and HAR messages, and early detection of visibility conditions and traffic flow characteristics may lead to reducing the number of accidents. The entire system, inc l uding camera manipulation, decoding equipmen t and camera site transmissions, can be operated from a central traffic management center In the recent past, several other innovative operational measures such as the PACE Program, Trucks at Rest in Fog (T ARIF), truck staging, truck metering, and truck convoying have been tested successfully." The California Highway Patrol (CHP) conducted field tests with a special enforcemen t unit called the PACE team between November 1991 and February 1992. The CHP used six units for patrol during weekday commuting hours along a 44-mile freeway segment when the vis i bility was limited to less than 200 feet. Over the 4-month evaluation period, a total of 144 hours of CHP time was provided at a total cost of $235,000. In this measure, the patrol units entered the freeway at staggered on-ramps on the test section with flashing lights, not allowing the vehicles to pass The officers selected the safest possible speed based on the prevailing visibility condition and paced the traffic at that speed before exiting the freeway and then re-entering in front of a different group of motorists to repeat the maneuver The CHP authorities concluded that the presence of law enforcement vehicles resulted in a speed 24 NCHRP Synthesis 22&. 2S 'Highway Accident Report on 1-40 Crashts, Arkansas. 29


reduction and a decrease in the number of collisions It was also noted that motorists began to call local media and traffic control centers to learn where the PACE team was working. Though the PACE has been tested successfully in California, it is difficult to conclude the efficiency of this rneasuxe from the limited information available, and it is also not clear how the officers were ab l e to control the rush hour traffic on multi-larie highways. Other operational measures like Trucks At Rest In Fog (TARIF) and truck staging involve encouraging truck drivers to delay or stop their trips during the fog periods voluntarily. For this pWJ!Ose, staging areas were constructed at each end of the test station to "hold" trucks during periods o f low visibility. Information on visibility, road conditi o ns, and control measures was also provided through pamphlets and brochures at staging areas. Truck metering and truck convoying also were tested successfully as possible countermeasures for poor visibility conditions. Various states in United States are engaged in analysis design, and in s tallation of several incident and motorist warning technologies. Leading advocate states are Alabama, Ark ansas, Georgia, New Mexico, Oregon, Tennessee, Idaho, New Jersey, South Carolina, Louisiana, California, and Utah. Alabama DOT is planning to install motorist system on a seven-mile sea bridge on I 1 0 in Mobile. This $3.4 million project will be equipped with an incident and motorist warning technology consisting of four new overhead variable message sign boards (two already exist), foux CCTV cameras, 14 surveillance type cameras, 12 variable speed signs Ail these com p onents will be integrated 'vith a control center that is already in place at the west end of bridge. VMSs are estimated to cost $941,000, speed signs about $24,000 each, CCTV cameras around $18,000 each, surveillance cameras are around $15,000 each. The operational costs are expected to be minimal, as most of the transmission equipment and control center with operators are already in place New Jersey has fog detection equipment connected to a light guidance system manufactured by 3M on I-287. This system includes a light guidance tube to illum i nate a 2000foot bridge on I 287 It is a delineation device that provides a visible line of light to guide drivers through road sections, especially a t night or during poor visibility conditions. This system consists of ultraviolet stabilized polycarbonate tubes with an optical l ighting film and follows the principle of "total internal reflection," which allows a low voltage source to illuminate a 100-foot section of connected tubes. Multiple sections are linked together to give 30


drivers a continuous illuminated delineation. The tube is activated automatically when the fog sensor detects that low visibility conditions are prevailing The color of the tube can be changed easily by changi n g the filter contain ed in the system. The tubes also cari be equipped to show different colors to drivers traveling in different directions Typically, for a 2,000-foot sectio n of a roadway, a light guidanc e systems cost about $25 ,000. These systems are being used in various states for a variety of purposes like steep curve negotiation, exit/entrance ramps, and construction work zones. South Carolina has had an incident and motorist waming system in operation for about six years. This system was designed to monitor conditions on the Cooper River Bridge, advise the motoring public of adverse conditions, and direct corrective actions. The system has four primary components: passive traffic control features, active traffic control features, weather detection equipment, and a surveillance system. The main objective of the system is to provide enhanced guidance for traffic in the bridge area. This is accomplished by using passive traffic control features like fixed signs, upgraded striping standards, and raised reflectorized pavement markers. The active part of the traffic control includes lighted pavement markers, street lighting control, and a VMS system with eight VMSs. All these components are connected to a control center with fiber optics and are computer driven. Eight surveillance systems consisting of color, pan, zoom, and tilt CCTV cameras also have been installed. The Conditions on the bridge fall into six classi ficati ons, and each condition has a programmed set of messages for the signs and directions to the diffe rent sections of the bridge. The Idaho Department of Transportation is in the process of field testing a motorist waming technology that it gets activated automatically, once the visibility sensors detect poor visibility conditions. The addition of two more variable message signs to the existing (two) chum-type changeable message signs is being contemplated. Tennessee also has a computerized incident and motorist warning system. This $4.5 million project encompasses a three-mile fog prone area of I-75 at the Hiwassee River crossing and eight-mile approaches on each side. Drivers are warned via one or more of the three HAR transmitters, 1 0 variabl e message signs, and 44 radar vehicle flow detectors Thresholds in changes of speed and/or flow automatically activate control messages on the VMSs. On-site communication between system components is provided by buried optical fiber cables, and the data is transmitted by microwave through two repeater sites to the control center 40 miles away from the project site. No fatal or property damage accidents have been observed since the installation of the warning system in April 1995 31


In Georgia, as previously mentioned, will be 1he one of the first fully-automated motorist warning systems in United States by the middle of 1997 This system is equipped with a network of 19 forward scatter fog sensors, 5 sets of highway -embedded speed monitoring loops to monitor traffic speed and volume, 4 changeable message signs, and several other weather instruments to measure precipitation, humidity, wind speed, and temperature. Two of the signs, which are 36 feet wide and 9 feet high, are installed over 'the traffic lanes. Two smaller signs, each measuring 16 feet wide by 9 feet high, are on the shoulder of the road. The latter could provide warnings to reduce speed or even provide detour instructions .. These sensor s, signs and speed -monitorin g loops will be connected to the traffic control center in Atlanta through telephone cables and transmission can also be upgraded with fiber optic cables in the future. The signs can be turned on manually by the local Cook County Sheriff's office in Adel. The variable message signs are expected to cost about $110,000 each. The weather station with precipitation, humidity, wind, and temperature measuring instruments, may cost in the range of $5,000 to $6,000. The entire project is estimated to cost just under $3 million. The Central San Joaquin Valley which encompasses the Fresno area in California, is equipped with several incident and motorist warning featwes like portable changeabl e message signs, higbway advisory radio, flow interruption technologies like CCTVs weather stations, and fog detectors."' It has four remote processor assemblies consisting of pavement sensors, small weather stations with visibility sensors, and a processing wlit in the Central Valley Traffic Operations Center (CVTOC). It also bas incident loop detectors installed at 27 locations and four CCTV monitoring stations to verify the operation of variable message signs. The CCTV system provides the visual information necessary to select appropriate CMS and HAR messages without delay. Several operational measures such as truck staging, truck metering and truck convoying have also been implemented. CAL TRANS is also in the process of installing another fog and motorist warning system in the Stockto n area. The proposed fog warning system will have field station/CMS (FS/CMS) sites, the substation (SIS) sites, and central computer with satellite terminal as its main components." The nine FS/CMS sites will include CMS's, fog sensors, and communication devices. The comJ!)JIJ)ication system will consist of direct burial twisted pair communication cables. A personal..,omputer-based central computer center bas also been planned for district headquarters in Stockton. This system detects reduced visibility 26 Final Report on 'Opmlion Fog" and NCHRP Report228. NCHRP R

conditions, and the vehicle detecton will detect the slowed/stopped traffic conditions without human input. In a recent study done by Louisiana Department of Transportation and Development on the fog-related accidents occurring on elevated roadwa y sections of Louisiana, several incident and motorist warning technologies have been suggested to counter poorvisibility conditions. This study recommended several countermeasures, covering more than 67 miles of elevated portions of roadways on 1-10, 1-55, 1-310, and US-190. They included installation of-variable message signs, use of advisory radios, installation of refiective raised pavement llliiikets, and the installation of barrier rail refiocton on all bridge sections without shoulders under study. The total cost of the project is estimated to be more than $2 million. The study recommended installatio n of seven variable message sign, expected to cost $700,000. It also recommended the installation of raised reflective pavement markers at a cost of $21,120 per mile and the use of barrier rail reflecton on the bridge sections at an estima ted cost of$2,000 per mile at a spacing of 105 feet.201 The study also stressed the need to strengthen the public awareness campaign to improve the driving habits during poor visibility conditions. An operational measure during heavy fog conditions is currently being applied along the 24-m.ile Lake Pontc hartraio Bridge. The right lane only is used in each direction with police units escorting vehicle p lat oons from the front, rear, and middle. Following an accident on Motorway 25 in conditions of patchy fog in 1984, the British Department of Transport installed an automated motorist and incident warning system to provide advanced information t o driven on the prevailing weather conditions. This system is equipped with several pre-programmed variable message signs. It is also noted that several other countries such as the Netherlands also have implemented a number of fog-related warning systems by using variable message signs, detection technologies, and surveillance technologies. Summarv It can be concluded that several advanced technologies should be considered to mitigate the adverse visibility conditions posed by fog. However the feasibility of advanced systems for automatic weather detection and motorist warning depends upon the characteristics o f each location such as topographical features, roadway geometry, prevailing speeds, and extent and 28 Louisiana DOT's study on FoaRelared Accidents 33


nature of recurring fog-related incidents. Benefits of inveS1ment versus effectiveness after ins1allation bave not been documente d in the literature or from discussions with project participants. For purpo ses of this evaluation report, syst em components and associated costs have been summarized in Table 12, as compiled from other projects previously referenced in this report. This serves as guidance towar d an "incremental approach" in technology application. In other words, if a parti cular area is found to be fog or fog crash prone in the future then the effectiveness o f a low-level technology application can be evaluated over time before significant investment is justified at a higher level (more elaborate combination of technologies). From bottom to top, this table can also be viewed as a hierarchy o f technology deployment for areas of recurring fog-related incidents. For now, the Tampa Bay Area should carefully monitor the results of the 1-75 fog detection and warning "prototype" system being dep loyed by the Georgia DOT before major inveS1ment in suc h systems is made Table 12 Fog D eteelio a and Motorist Waroing Syste m ComponeotS SYSTEM COMPONENT ESTIMATED COST Variable Message Siges $75,000-$200,000 e ach V ariable Speed Signs $15,000-$24,000 each CCTV/Survelllan< Cameras $15,000-$18,000 each Jnwgrated Weather StationsOO SS,000-$6,000 each F og Sensors SS, 000-$8 ,000 each Raised Pavement Markets $21,120 per mile <'> (a) includes precipitation, humidity temperature, and wind measuring instruments. (b) assumes 5 fl spacing a l ong edge lines and 10 fl spacing atona centerline for 2 lanes. c) all costs have summarized from previously referenced reports, 1992.-1996. It is believed that technologies probably cannot provid e effective solutions if problematic locations are d ispersed and scattered. According to the Louisiana D e partmen t of Transportation and Development, "The state Can provide detection, warning, and guidance technologies, but 34


much of the responsibility must be placed on the motorists to adjust their driving habits to the environmental conditions. Without the motorists changing their driving habits during times of reduced visibility, these accidents will continue resulting in some catastrophic accidents at some time."29 VII. Driver Education and Awareness Techniques Driver perceptions and responses are important during conditions of poor visibility b e cause poor visibility conditions complicate driving tasks. Driver problems in fog include: restricted visibility; gpeed election beyond available visibility; over response to changes in vehicle speeds; sudden lane changing; and lack of knowledge on poor visibility crashes. The National Transportatio n Safety Board (NTSB) has noted that, in many chain type fog-related crashes, various investigating agencies have attributed the cause to "Driver error." An example of one of the most-observed driver errors related to fog accidents have been "driving too fast for conditions." There are, of course, examples in which drivers travel at 60 mph under zero visibility conditions fully aware of the hazards and willing to assume risks. This is not driver error but rather a disregard of rights and safety of others in the use of a highway. An example of a driver error is simply stopping on the traveled portion of the highway, thus creating the first link in a chain-type accident. Another example would be passing another vehicle without assured clear distance ahead. One of the most serious problems concerning the drivers in limited visibility is choosing a safe speed. The NTSB determined that the one main cause o f poor visibility crashes is the non uniform regponse of drivers and concluded that drivers tend to operate at varying gpeeds. Several highway accident reports pointed out that, as the drivers approach and enter the fog area, they reac t in different ways. Some drivers may reduce their gpeed, some may turn on headlights and/or warning flashers, and others either may adopt a wait-and-see attitude before entering the fog area. Although the travelers could see the fog surrounding the highways, they may perceive risks differently and pursue their journey, lacking gpecific behavioral guidance.30 Very few studies have been done on driver behavior during poor visibility conditions. A 29 Louisiana OOID's srudy on fog. 30 "Highway Accident Repon on 1 Crashes," Arkansas. 35


1967 study concluded that in poor visibility conditions mean and 85th percentile speeds would reduce by 5-8 mph, but it also observed that some drivers proceed at speeds higher than posted speeds. The posted speeds were observed to have significant impact on speed variations, however, posted speeds less than 35 to 40 mph had little impact on the speed reduction.31 Another study done in Oregon indicated that lower visibility conditions result in lower speeds, and this study also emphasized the importance of signing in advance of the fog and also in the fog area. A questionnaire survey concerned with driving habits performed as part of the survey indicated that, 46 percent of the drivers preferred to follow another vehicle in fog, 29 percent preferred to follow pavement markings, and 5 percent of the drivers indicated their preference to pull off the road and stop theirtrip.32 A vehicle speed analysis done in Idaho has been successful in answering two critical questions concerning driver-behavior in bad weather. The evaluation study done to test the Idaho storm warning system concluded that drivers indeed respond to poor visibility conditions by reducing their speeds, and the average drop in speed observed was about I 0 mph. Another important observation from this study relates to non-uniform driving speeds It concluded that the variability in individual v ehicle speeds will be higher in poor visibility conditions when compared with normal conditions. These findings validate the observations made by NTSB on the aspect of non-uniform driving bebavior.33 The problem of non-uniform driver behavior requires several measures to be taken to ensure that guidance for driving in limited visibility conditions be uniform and complete. The introduction of a warning system ahead of the initiation of a response serves to increase the time available for reaction. Previous studies revealed that drivers react 1.35 times faster to the anticipated stimulus than the unexpected stimulus (0.54 to 0 .73 seconds) Another study showed that a warning signal with an optimal lead time of 200 milli-seconds could reduce reaction time by 50 milli seconds." Though these studies s i gnify the advantages of the presence 31 NCHRI' Synthesis 228. n 'Speed Advisocy Information for R

of a warning system before a stimulus and response, there is no comprehensive evidence available to suggest that the provision of advance warning systems like speed signs, variable message signs, and highway advisory radios consistently lead to speed reductions. A 1979 study do n e in Oregon indicated that the installation of variable message signs may not result in speed reductions.36 Another Virginia study experimenting with pavement insert lights concluded that the improved delineation may indeed increase the potential for accidents with the increase in night time speeds. The NTSB also found that of the drivers involved in crashes due to fog lacked knowledge about whether they should leave or stay in their stopped vehicles. Unfortunately, none of the states outside of California associated with poor visibility crashes attempted to educate drivers in this area." In addition to electronically operated warning systems, extensive public awareness programs consisting of review and updating of remedial training material and driver license material are important in mitigating the poor visibility problems. Several highway accident reports previously referenced indicate the driver's tack of caution as a reason for poor visibility accidents. However, the drivers involved in these crashes cited their lack of knowledge and lack of training in evas i ve procedures during fog conditions Such low awareness problems can be solved largely with some well-coordinated public awareness campaigns. However, it is found that, among various states affected with poor visibility problems, California is the only state. that is spending time and resources on public awareness campaigns. An example of one of California's public information brochures is noted in Figure 4. NCHRP Synthesis 228. 37 "Highway Aocident Report on Corona Crashes, California. 37


FJcure 4 Califomia Public Broebure for Fog FOG SEASON ;.;,.;...o.;.,.,,.._c$_' . .: .......... 'tn"tol_... rwa,_.-.' Mt, ..... ..... "'UI)!...G 4oo)ll PRONE' AREAS r; 38 .... ,.,..,.,. .. ._ QIIO.Orilr ........... ...,.._ ..... -, -...o-w ... ......


Public awareness strategies adopted by California include several elements like multi lingual pamphlets, brochures, posters, and pul;>lic service announcements (PSAs). A local Tampa Bay area example of a public awareness strategy is exemplified by the recent newspaper supplement entitled 1997 Hurricane Survival Guide, sponsored by the Tampa Tribune, Radio Shack, and News Channel 8. This guide includes a storm tracking map, storm classifications, a survival checklist, and emergency management numbers. A similar guide could be developed "for driving in fog and distributed during or just prior to the December-February fog season in Tampa Bay Awareness programs in California have been designed to include information on general fog formation characteristics, fog and fog prone areas, and tips for driving in the . conditions of poor visibility. These brochures and pamphlets were distributed through varioils agencies like highway patrols, trucking associations, truck stops, truck terminals, civic organizations, major employers, media outlets, automobile associations, insurance companies, local citizen groups, and special safety programs. Posters are designed for display at rest areas and truck stops to acquaint motorists with the measures to be taken in limited visibility conditions. Presentations have also been made to community groups and to trucking company officials and drivers. The radio PSAs used sound effects like fog horns and police sirens to get the attention of listeners News releases and press-conferences involving news papers, radio, and TV are the other media employed for carrying out public awareness programs. CAL 1RANS also made significant attempts to elicit the public perceptions and responses of the usefulness of the countermeasures implemented. By publishing a questionnaire in local newspapers, they obtained inputs from the traveling public on various countermeasures implemented such as fog pamphlets, VMSs, HA.Rs, T ARIF, and truck staging. The results from the survey indicated a favor able response rate of 80 to 92 percent, which is a clear indication of success of the countermeasures, however, fog pamphlets received only 53 percent favorable response rate." This has been attributed to the fact that the questionnaire published in newspapers was available to the residents of the entire valley, all of whom were not the targets of the fog pamphlet. Drivers who do decide to venture out into heavy fog should be individually responsible for taking the necessary precautions to avoid collisions. As a start for public awareness, based on general information provided by the American Automobile Association and excerpts from a December 31, 1996, Tampa Tribune editorial, th e following safe driving tips in fog are offered. NCIIRP Synthesis 228. 39


Table 13 Safe DriYing Tips in Fog How to Drive in Fog Consider

vm. Conclusions and Recommendations Between 1987 and 1995, fog-related crashes represented 0.32 percent of total roadway crashes in the state of Florida. Within Hillsborough and Pinellas counties, fog-related crashes represented about the same proportion of total crashes (0.30) during the same period. Fog related crashes for this period resulted in 300 fatalities statewide, 29 of which occurred in Hillsborough and Pinellas counties. Nationally, in 1994, the U.S . average for fog-related (weather condition only) fatal crashes only was 1.6 percent of total fatal crashes (2.2 percent for Florida in the same year). Based on thjs report. jt has been determined that there are no particular fog-prone or fog-crash-pro n e areas jn the Tampa Bav area. However, there is a fog season that occurs primarily between December and February. These are the months when heavy fog is typically reported for at least 3-4 days each month. The crash rate for fog-related crashes has been above the statewide average in Hillsborough County and below the statewide average in Pinellas County. Additionally, over the last I 0 years, more fog-related crashes have been reported in Hillsborough County than any other county in Florida. Leading advocate states in the installation of fog detection and motorist warning systems include Alabama, Arkansas, Georgia, New Mexico, Tennessee, Idaho, New Jersey, South Carolina, Louisiana, Oregon, Utah, and California. Several of these states have deployed $2$3 million weather detection/motorist warning systems along specific travel corridors, but the benefits of these systems have yet to be documented. Common in all of the individual state reports examined was the recommendation to improve driver awareness for driving in fog (along with the technology applications to poor visibility mitigation). However, only California has actually invested time and funding toward a focused public awareness campaign, which has received positive public feedback. The National Transportation Safety Board has determined that the single cause of poor visibility crashes is non-uniform response of the drivers. Further, a recently completed statewide fog crash evaluation study done in Louisiana concluded that "the state could provide warning and guidance technologies, but much of the responsibility for safety ultimately must still be placed on the motorists to adjust their driving hab its during times of reduced visibility." In order to reduce fog-related crashes in an area with seasonal but scattered fog-prone and fog-crash-prone areas, a major investment in detection and warning technology would not be warranted at this time. Some minimal applications of low-level visibility enhancement and warning (raised pavement markers and/or variable speed signs) could be evaluated on an experimental basis for effectiveness in the most heavily-traveled corridors where fog crashes 41


have occurred, only as uncommitted funding becomes available. A driver awareness program would be the most cost-effective countermeasure at the present time, given the aforementioned findings. This report recommends that a very focused driver awareness campaign be initiated just prior to and during the fog season of December-February. Given the characteristics of fog related crashes that have occurred over the last decade, it appears that this awareness campaign should be aimed at: Hillsborough more than Pinellas County residents, passenger car owners, between the ages of20-29, driving during the morning commute how:s, on local and county roads in rural locations. Public service announcements, simple brochures describing driving tips in fog (see Table 13) and fog formation characteristics, and enhanced traffic reporting on radio and television highlighting current and historical fog information during the "fog season" would be most appropriate. Slowing down or delaying trip altogether would be of the more prominent messages to the public during heavy fog conditions. As they have for the "hurricane season," the News Channel 8 weather team could be prominent in the PSAs. A monitoring aspect of the driver awareness campaign should also be included to determine effectiveness and trigger possible future detection/warning technology applications. 42


BffiLIOGRAPHY Adverse Weather, Reduced Visibility Conditions, and Road Safety (a report on driving in reduced visibility conditions due to adverse weather), Organization for E conomic Co-operation and Development, 1976. Brisbane, G.J.B., "Driver Response to Fog Conditions: An Intelligent Pacific Rim TransTech Conference Proceedings, Volume I, New York, NY, 1993. Dahlinger, Don., "Fog Warning System Provides a Safety Net for Motorists Public Works, Vol. 126, No. 13, 1995. "Disaster on the Sunshine Skyway Tampa Tribune, December 31, 1996. "DOT considers Fog-Warning Systems, St Pete Times, January 1, 1997. "Driving in Bad Weather," video by American Automobile Foun dation For Traffic Safety, 1994 . "Evaluation ofM25 Automat ic Fog Warning and Assessment Systems," Acer Consultants Ltd., Guilford, UK, 1993 "Examination of Reduced Visibility Crashes and Potential IVHS Countelllleasures," U.S. Dept. of Transportation, 1995. "Fog Clouds Picture in Tampa Bay Area," Tampa Tribune, January 4, 1997. "Fog shuts Skyway bridge duringmomingrush hour," St. Pete Times, December 31, 1996. George, L.E., Hofstetter, D.K., and Wagner, D .R.. ''Variable Message Fog Hazard Warning Signs to Control Vehicle Operating Characteristics, Report No. FHW AIOR-79/3, Oregon DOT, FHWA, Washington D.C., 1979. Heiss, William H., "Highway Fog: Visibility Measures and Guidance Systems," NCHRP 171. "Highway Accident Report on Multiple Vehicle Collision During Fog Near Milepost 118 on Interstate 40, Menifee Arkansas, January9, 1995,"National Transportation Safety Board, Washington D.C., 1995. "Highway Accident Report on Multiple-Vehicle Collisions During Limited Visibility (FOG) on Interstate 75 Near Calhoun, Tennessee, December 11, 1990," National Transportation Safety Board, Washington, D.C., 1992 "Highway Accident Report on MultipleVehicle Collisions and Fire under Limited Visibility Conditions, Interstate Route 75 at Ocala, Florida February 28, National Transporta tion Saf ety Board, Washington, D.C., 1983. 43


"Highway Acddent Report on Multiple Vehicle Collissions Under Fog Conditions and Fire, New Jersey Turnpike, November, 29,1969," NTSB, Washington, D.C., 1969 Hinson, Joe, "Common Sense on Roadways," Tampa Tribzme, January 5, 1997. "How to Drive, A textbook for Busy Adults,'' American Automobile Association, 1993. Idaho Storm Warning System IVHS Operational Test, Plan, CH2M Hill, Boise, Idaho, 1993. "Intelligent Transportation Systems: 1-75 Fog Detection/Warning System," USDOT, FHW A. "Interim Report on Reduced Visibility (fog) Study," California High Transportation Agency, 1965. "1-526 Cooper River Bridge Surveillance and Response Plan," South Carolina Department of Highways and Public Transportation, June, 1992. Moore, R.L. "Fog and Road Traffic," TRRL Report: LR 446, Transport and Road Research Laboratory, 1972. McPhee, Ian J et. al., "The Development of a Fog Potential index for English Motorways," University of Birmingham, 1995. "Pileups on Foggy Bridge Stall Bowl Trips," St. Pete Times, January 1, 1997. Report on "Highway Visibil ity Conference," November 6-7, 1996, Huntsville, Alabama. Sayed, Tarek, and Zein, Sany, "General Review of Advanced Tec hnologies," G.D. Hamilton Associates Consulting Ltd., British Columbia, Canada, 1995. Shannon, Patrick, et a!., "Idaho Storm W aming System ITS Operational Test Phase 1 Interim Report," Idaho Department of Transportation, 1997. Shepard, Frank D., "Reduced Visibility Due to Fog on the Highway NCHRP Synthesis 228, Transportation Research Board, 1996. "Special Public Hearing: Fog Accidents on Limited Access Highways," Nati onal Transportation Safety Board. Washington, D C, 1992. "Special Study: Reduced Visibility (fog) Accidents on Limited-Access Highways," Report No. NTSB HSS-72.4, National Transportation Safety Beard, Washington D.C, 1972. Strahler, Arthur N., Physical Geography, Third Edition, John Wiley and Sons, Inc, 1968. "Touch Weather Wisdom,'' The Wisconsin Weather Team, News Cbannel4, Jan. 7, 1997 and May 28, 1997. "Troopers, Fog Shut Down Sunshine Skyway;" Tampa 1Hbzme, December 31, 1996. 44


User's Manual on "Caltrans Meteorological System," Report 171, Qualimetrics, Inc, 1997. Wagner, D.R., and Hofstetter, D.K., "Speed Advisory Information for Reduced Visibility Conditions," Report J::lo. FHW AIRD-78/32, Oregon DOT, FHW A, 1978. Walter, J.D., "Strategies to Reduce Multi-vehicle Collisions During Limited Visibility Conditions,'' Caltrans, September, 1992. White, M.E., some Aspects ofMotorway Traffic Behavior in Fog," TRRL Laboratory Report 958 0305-1293, Highway Traffic Division, Traffic Engineering Dept., Transport and Road Research Laboratory, 1981. 45


APPENDIX (Crash Rate Tables, by County, by year) 46


Crash Rates by County 1987 No. Rank :ill: 72 I , Lee Matlqn Liberty St. L u cie c Johns 11 4 z r o


Fog.ftel ate d Crash Rate s by County 1989


Fog-Related Crash Rates by County 1990 30136(/612 28


Fog-Rolated Crash Rates by County 1930


Crash Rares by Co.cty 1991 ..E!!!l } :fi: ,-----'---' Jli 31 p! ll '--:---,___ ;--:-----


0 665St2 3:181UIZ1 0 60


FozR elatecl Crash Rate:$ b)' Couty 19'9-4 =ii Rank ;-'----'-,--'-c-'-'-'-'-'-FHiff. c-,_ OOfiOf 21 32 PUoo '-OuVSl '-'-41 15 42 14 43 P a lm 50 1 :--- ''-,__ ,_ ,___ '5 Dado 1i :-ii '--0 .. 0 .. -0 2830749 0 .. Statowldo 485 15


Foc-Rtlattd Cnd1 Ra.us h)' Cca.at)' "" 0 ::::::: 0


Foc-Rtlattd Cnd1 Ra.us h)' Cca.at)' "" 0 ::::::: 0


Download Options

Choose Size
Choose file type
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.